The Making of Economic Policy in Africa - World...

8775-hmgl Economic Development Institute

,I;,# of The World Bank8775

The Making of EconomicPolicy in Africa

Ravi Gulhati

EDI SEMINAR SERIES

FILE COPY

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

EDI SEMINAR SERIES

The Making of Economic Policy

in Africa

Ravi Gulhati

The World BankWashington, D.C.

Copyright © 1990The International Bank for Reconstruction and Development / THE WORLD BANK1818 H Street, N.W.Washington, D.C. 20433, U.S.A.

All rights reservedManufactured in the United States of AmericaFirst printing April 1990

The Economic Development Institute (EDI) was established by the World Bank in 1955 to trainofficials concerned with development planning, policymaking, investment analysis, and project im-plementation in member developingcountries. At present the substance of theEDI' s work emphasizesmacroeconomic and sectoral economic policy analysis. Through a variety of courses, seminars, andworkshops, most of which are given overseas in cooperation with local institutions, the EDI seeks tosharpen analytical skills used in policy analysis and to broaden understanding of the experience ofindividual countries with economic development. In addition to furthering the EDI's pedagogical ob-jectives, Policy Seminars provide forums for policymakers, academics, and Bank staff to exchangeviews on current development issues, proposals, and practices. Although the EDI's publications aredesigned to support its training activities, many are of interest to a much broader audience. EDI ma-terials, including any findings, interpretations, and conclusions, are entirely those of the authors andshould notbe attributed in any manner to the World Bank, to its affiliated organizations, or to membersof its Board of Executive Directors or the countries theyrepresent. The World Bank does not guaranteethe accuracy of the data included in this publication and accepts no responsibility whatsoever for anyconsequences of their use.

Because of the informality of this series and to make the publication available with the least possibledelay, the typescript has notbeen prepared and edited as fully as wouldbe the case with a more formaldocument, and the World Bank accepts no responsibility for errors.

The material in this publication is copyrighted. Requests for permission to reproduce portions of itshould be sent to Director, Publications Department, at the address shown in the copyright noticeabove. The World Bank encourages dissemination of its work and will normally give permissionpromptly and, when the reproduction is for noncommercial purposes, withoutaskinga fee. Permissionto photocopy portions for classroom use is not required, though notification of such use having beenmade will be appreciated.

The backlist of publications by the World Bank is shown in the annual Index of Publications, whichis available from Publications Sales Unit, The World Bank, 1818 H Street, N.W., Washington, D.C.20433, U.S.A., or from Publications, Banque mondiale, 66, avenue d'Iena, 75116 Paris, France.

At the time of writing, Ravi Gulhati was senior adviser in the Economic Development Institute of theWorld Bank. He is now a practicing consultant and an academic.

Library of Congress Cataloging-in-Publication Data

Gulhati, Ravi.The making of economic policy in Africa / Ravi Gulhati.

p. cm. - (EDI seminar series)Includes bibliographical references.ISBN 0-8213-1341-X1. Africa, Sub-Saharan-Economic policy. 2. Africa, Sub-Saharan-

-Politics and government-1960- 1. Title. 11. Series.HC800.G858 1989338.967-dc2O 89-22638

CIP

EDI Catalog No. 420 /088

CONTENTS

Abbreviations v

Preface vii

1. Introduction 1

2. Economic Policies and Their Consequences 7

The Main Facts 7The Role of Shocks 12The Role of Policy Factors 14Summary 21

3. Impact of Politics on Economic Policy 25Major Political Trends 26Policy Circle 29Policy Climate 29Political Parameters and Policy Content 30Policy Implementation 33

4. Policy Experience in the 1980s 37Macroeconomic Changes 37Hostile International Environment 40Different Policy Responses 41Conclusion 57

5. Three Case Histories of Reform: Mauritius,Malawi, and Zambia 59

Brief Profiles 59Extent of Disequilibrium 61The Policy Turnaround and Its Impact 66Conclusion 74

6. Future Policy Reform 77Improving the Capacity and Willingness of a Country to Undertake Reforms 77Improving the Technical Design of Reform Packages 81Improving the International Framework 91Conclusion 97

References 101

List of Tables2.1 GDP and Agricultural and Manufacturing Value Added, 1967-73 and 1973-79 82.2 Gross Domestic Expenditure, Total Investment, and Consumption,

1967-73 and 1973-79 92.3 Exports and Imports, 1967-73 and 1973-79 10

iii

iv The Making of Economic Policy in Africa

2.4 Indices of External Financial Disequilibrium, 1973-79 102.5 Indices of Internal Financial Disequilibrium, 1973-79 112.6 Extent of Economic and Financial Crisis, 1970s 122.7 Magnitude of Exogenous Shocks 132.8 Characteristics of the Public Sector, 1970s 152.9 Nominal Protection of Taxation of Agriculture, 1976-80 and 1978-80 18

2.10 Indicators of Efficiency of Manufacturing 222.11 Economic Profile of Twelve Sample Countries, 1970s 234.1 Indices of Economic Change, 1979-86 384.2 Indices of External Financial Disequilibrium, 1980-86 394.3 Indices of Internal Financial Disequilibrium, 1980-86 404.4 Impact of Exogenous Shocks During the 1980s 414,5 Intensity of Policy Responses, 1980-86 424.6 Uganda: Chronology of Selected Policy Events, 1980-86 434.7 Zaire: Chronology of Selected Policy Events, 1979-86 444.8 Madagascar: Chronology of Selected Policy Events, 1980-86 464.9 Sudan: Chronology of Selected Policy Events, 1979-86 48

4.10 Somalia: Chronology of Selected Policy Events, 1980-86 504.11 Tanzania: Chronology of Selected Policy Events, 1980-86 514.12 Zimbabwe: Chronology of Selected Policy Events, 1981-86 534.13 Kenya: Chronology of Selected Policy Events, 1978-86 555.1 Selected Economic and Social Indicators, 1970-80 605.2 Mauritius: Chronology of Selected Policy Events, 1978-85 675.3 Malawi: Chronology of Selected Policy Events, 1979-86 695.4 Zambia: Chronology of Selected Policy Events, 1981-86 71

ABBREVIATIONS

ADMARC Agricultural Development and Marketing Corporation (Malawi)BDP Botswana Democratic PartyCAM Comit6 d'Action Musulman (Mauritius)COLAS Cost of living adjustment (System in Mauritius)COPWE Commnission to Organize the Party of the Working People of EthiopiaCPs Contract Plans negotiated between government and individual state

enterprisesDC Development Certificate (Mauritius)DRC Domestic resource cost of earning or saving foreign exchangeDWC Development Works Corporation (Mauritius)EACM East African Common MarketECA Economic Commission for AfricaEEC European Economic CommunityEFF Extended Fund Facility of the International Monetary FundEPZ Export Processing Zone (Mauritius)EROP Effective rate of protection from importsESAF Enhanced Structural Adjustment Facility (IMF)FAO Food and Agriculture Organization of the United NationsGDE Gross domestic expenditure (consumption plus investment)GDP Gross domestic product (value added in all activities)GDY Gross domestic income (GDP corrected for changes in terms of trade)IBRD International Bank for Reconstruction and DevelopmentIDA International Development AssociationILO International Labor Organization of the United NationsIMF International Monetary FundLIBOR London interbank offered rateMLP Mauritius Labour PartyMMM Mouvement Militant Mauricien (Mauritius)MCP Malawi Congress PartyMSM Mouvement Socialiste Militant (Mauritius)MUZ Mineworker's Union of ZambiaNAMBOARD The National Agricultural Marketing Board (Zambia)NCPB National Cereals and Produce Board (Kenya)ODA Official development assistanceOECD Organization for Economic Cooperation and DevelopmentPHL Press Holding Ltd. (Malawi)PCU Provincial Cooperative Union (Zambia)PMSD Parti Mauricien Social et Democrate (Mauritius)PSM Parti Socialiste Mauricien (Mauritius)QRs Quantitative restrictions on importsROs Remuneration orders (system for regulating labor market in Mauritius)

v

vi The Making of Economic Policy in Africa

SAF Structural Adjustment Facility of the International Monetary FundSAL Structural Adjustment Loan offered by the IBRD or IDASCP Sudanese Communist PartySECAL Sectoral Adjustment Loans (IBRD and IDA)SINPA Societe Int6ret Nationale Pour l'Agriculture (Madagascar)SPLM Sudanese People's Liberation MovementSSA Sub-Saharan AfricaUNCTAD United Nations Conference on Trade and DevelopmentUTNDP United Nations Development ProgrammeUNO United Nations OrganizationZANU Zimbabwe African National UnionZAPU Zimbabwe People's UnionZCCM Zambia Consolidated Copper Mines

PREFACE

I started to work on economic issues in Sub-Saharan Africa in 1977. As the World Bank'schief economist for the Eastern and Southern Africa Region, I had a professional interest intwenty countries, including the dozen analyzed in this book. During the period 1977 to 1986, 1often visited these countries, read a great deal of the economic literature about them, andhad numerous opportunities to talk with government officials and people working there inthe private sector. When the World Bank started its policy-based lending operations in 1980(up until then loans typically financed well-defined projects), my work began to focus on keyissues of macro or sectoral policy. I was personally involved in negotiating policy agreementswith many member governments. It was during this period in which analytical work oneconomic policies became a major preoccupation that I began to fathom the limitations ofdevelopment economics and to feel the need for a broader framework for analysis.

Writing this book has led me to reflect on this experience with the benefit of hindsight.It has also made me systematically explore the political and historical aspects of policychange for the first time. This has been something of an education for me, and it hasinfluenced my thinking. Readers of this book who have known my earlier writing willundoubtedly find evidence of a new orientation.

Some readers may be surprised by the critical tone in certain passages. Indeed, I have notshied away from a candid discussion of economic policy in Africa because I believe that it isonly by avoiding diplomatic double-talk that a contribution can be made to an improvedunderstanding of this subject. It is important, however, to put the African scene in properperspective. The making and administering of economic policy in all developing countries(and in the industrialized world) reflect a combination of technocratic, political,organizational, and personal considerations. In this broad sense, political in-fighting,ministerial turf battles, and patron-client relations are not phenomena peculiar to Africa.Newspaper readers everywhere are aware of these dimensions of the policy universe. Whatis perhaps distinctive about Sub-Saharan countries in the first quarter-century of theirpolitical independence is that the balance between the meritocratic, objective, andprofessional considerations on the one hand and parochial interests on the other is tiltedheavily toward the latter. And even this conclusion will not surprise historians who haveanalyzed the very slow evolution over many centuries of civic norms, public ethics, andrepresentative government in Europe and elsewhere.

In January 1989 the Economic Development Institute of the World Bank published casestudies by me on Zambia and Malawi-and the following year a third on Mauritius. Allthree discuss policy change in response to internal developments and external eventsaffecting these countries. In this book I summarize this earlier work and extend my analysis,as explained in Chapter 1. Notwithstanding many connections with the EDI and the WorldBank, the entire project is my own. The impact of politics and sociology on economic policy isa sensitive area in which the Bank has no official views. Judgments made in this book arepersonal, and I alone am responsible for them.

Raj Nallari painstakingly assembled research materials and participated in theanalytical work. Sofia Mendoza also assisted in many ways too numerous to spell out.Without the valuable contribution of these two colleagues, it would not have been possiblefor me to write this book. All of us would like to express our gratitude to Yukoshimamoto ofthe Africa Information Service Center at the World Bank for the help she provided.

I also extend my thanks to Christopher Willoughby, the director of EDI, for stimulatingand supporting this exercise. His intellectual curiosity and courage to launch new initiativesare largely responsible for this undertaking. The entire manuscript was reviewed by him aswell as by Steve O'Brien, Robert Armstrong, Timothy King, Cadman Mills, and Bimal Jalan.These readers (and the much larger group who read individual chapters) inspired me torethink and rewrite.

Ravi Gulhati

vii

1

INTRODUCTION

Efforts of governments in Sub-Saharan Africa (SSA) to improve their economic policiesduring the 1980s sparked intense debates among politicians, officials, and scholars. Thesedebates are bedeviled by the absence of a clear framework and a good definition of issues.This chapter provides a few illustrations of these exchanges. My main aim is to examine themajor policy developments in SSA. The study analyzes the substance of these changes andalso the underlying governmental processes. Developments in these economic policies from1980 to 1986 are assessed in relation to the history of the postindependence period. Ananalytical scheme for thinking about economic policy is presented. Armed with such ascheme, readers should be better able to sort out the puzzling controversies surrounding policychanges in Sub-Saharan Africa and to gain a more meaningful understanding of key issues.

Let me illustrate one aspect of the recent controversy. At a conference on the "Crisis andChallenge of African Development" in September 1985 in Philadelphia, I was struck by thegulf separating the views of American political scientists specializing in African studies andthose of Washington bureaucrats in international organizations or the U.S. government.Although the approach of the conference was "overtly interdisciplinary andtransideological" (Glickman 1988, p. x), I left the meeting with some disquiet. Apparently,some scholars at the conference viewed the reforms being urged on African governments byforeign aid donors as an attack on the African state. To quote Glickman,

Austerity undermines authority in circumstances where legitimacy in great partdepends on distributive capabilities. A reduction in the overhead authority of the state,especially in the midst of the attempt to introduce austerity reforms, opens up newopportunities for ethnic and class conflicts. Suppression of conflict in turn risks relianceon force alone (Glickman 1988, p. 230).

I became determined to delve further into the views of political scientists whoseperspective was so alien to those of professional economists. Recent seminars organized bythe Economic Development Institute of the World Bank for senior policy officials from Sub-Saharan Africa reinforced the view that political issues in these countries were a majorimpediment to economic policy reforms. To promote greater understanding of these issues, Ipersuaded EDI to convene at the end of 1986 two workshops in Dalhousie, Canada, andWashington, D.C., with North American and African political scientists. While debate wasvigorous and many insights were gained, the absence of a clear framework for the treatmentof political economy issues was evident (Gulhati 1988, p. 5). Many scholars were extremelycritical of Bank-Fund proposals, but attempts to outline alternative approaches to Africa'seconomic problems did not get very far.

Yet another illustration of the controversy over economic reforms in Sub-Saharan Africa isthe sharp contrast between the views of the World Bank and those of the EconomicCommission for Africa (ECA). Ever since the World Bank (1981) published AcceleratedDevelopment in Sub-Saharan Africa: An Agenda for Action (popularly called the BergReport) that outlined the required economic reforms in the region, ECA has fired manysalvos. The World Bank, in turn, has produced a series of reports aimed at elaborating ormodifying its earlier analysis. Controversy between the two organizations recentlyresurfaced. The World Bank and the United Nations Development Programme (1989)published a report that identified "emerging signs of economic improvement since the mid-1980s" (p. iii); it also claimed that "programs of economic reform and adjustment have

1

2 The Making of Economic Policy in Africa

helped African countries begin to improve their economic performance" (p. iii). Meanwhile,a recent ECA document (undated, p. 16) asserted that despite many stabilization andstructural adjustment programs sponsored by the IMF and the World Bank during the pastdecade, "the crisis remained unabated. Many African economies moved from stagnation todeclining growth; food deficits reached alarming proportions; unemployment mounted;underutilization of industrial capacity became widespread; and environmental degradationthreatened the very survival of the African people." The report went on to say that "giventhe structure of Africa's socio-economic structures, and the region's development objectives,the orthodox approaches to stabilization and adjustment are inappropriate for bringingabout economic recovery and socio-economic transformation" (ECA undated, p. 26).

The intellectual divide on reform issues in Sub-Saharan Africa is neither on a North-South axis nor on an economist versus political scientist basis. The Berg Report, for example,was vehemently criticized by economists and other social scientists in Sussex, England, atthe Institute of Development Studies (1983). Once again the policy-based operations of theIMF and the World Bank were subjected to serious scrutiny by European academics(economists and others) and by government officials at a seminar organized by theScandinavian Institute of African Studies (Havnevik 1987). Criticism focused particularly onthe alleged neoclassical philosophy of the IMF and the World Bank, the lack of effectivecollaboration between them, and their neglect of political realities in determiningconditionality.

Judging from EDI seminars (1981-89) for high-level officials from Sub-Saharan Africa (atwhich all the main stabilization and structural adjustment policies were discussed), I haveconcluded that support for or opposition to IMF-Bank assessments and recipes variesenormously from issue to issue. For example, the role of exogenous shocks and policy issues inexplaining the economnic crisis tends to be controversial, yet there is considerable agreementon the need for fiscal and monetary discipline, for improving the efficiency of the publicsector, and for redressing the neglect of agriculture. Much more controversial are IMF-Bankpositions on exchange rates, interest rates, privatization, and import liberalization. I havealso noted that the prevailing sentiment of SSA officials on particular policies varies overtime. For example, there was considerable skepticism about an active exchange rate policywhen the EDI seminars started in 1981, but by 1987 there had developed a much morebalanced appreciation of what such an instrument could or could not do.

At a much broader level, many African participants in these seminars in the early 1980sfavored the New International Economic Order, which involved the reform of globalinstitutions and the underlying economic policies of major powers. More recently, however,much less has been heard about these grand issues. Instead, attention has focused onimprovements that could be made by African governments and international agencies withinthe existing international system. The option of African countries "delinking" from theinternational system, advocated by the "dependency school," was seldom mentioned duringthese discussions. Whether EDI seminars provide a firm basis for making generalobservations about policy attitudes of SSA officials is an open question. It is conceivablethat these seminars have a special "atmosphere" absent in other forums. African officialsmeeting under the auspices of the ECA or working in their own governments probably adoptpolicy positions that reflect the distinctive characteristics of these institutional settings.Such behavior is fairly common everywhere. Where we stand on policy questions dependsquite a lot on where we sit.

Too much of the writing on the economic crisis and policy reforms in Sub-Saharan Africais pitched at the regional level. This obscures the considerable diversity within the regionand adds to the difficulty of understanding what is being said. Recently some country casestudies have appeared that are very welcome, and I refer to them later on. In this book Iwill focus on a sample of twelve countries in Eastern, Central, and Southern Africa:Ethiopia, Kenya, Madagascar, Malawi, Mauritius, Somalia, Sudan, Tanzania, Uganda,Zaire, Zambia, and Zimbabwe. This is a diverse group. Some countries have relatively largepopulations (Ethiopia, Zaire), and one is a micro state (Mauritius). Most of the countries are

Introduction 3

primarily agricultural, but Zambia has a large mining sector and Zimbabwe a largemanufacturing sector. Per capita income in 1986 ranged from US$1,200 in Mauritius to US$120in Ethiopia. The sample countries are also politically diverse. Mauritius is the onlydemocracy, but the remaining authoritarian regimes differ significantly. The economic crisishas affected some countries only moderately and others severely. Certain governmentsresponded quickly and forcefully to the economic difficulties by carrying out economicreforms. Others experienced protracted delays in adopting reforms and encountered severeimplementation obstacles. Although the sample is not a representative one in the statisticalsense, it does include many types of countries, many varieties of policy problems, and manykinds of policy responses. The sample does not include any country from West Africa, but it isnot clear that countries situated in the western part of the continent are significantlydifferent from those in the sample with respect to issues discussed here.

I have selected Zambia, Malawi, and Mauritius for intensive examination because thesethree countries span a spectrum stretching from aborted reforms at one end to seeminglysuccessful reform at the other. In Zambia the government took a very long time to develop areform effort commensurate with the extreme gravity of its problems. The effort was madeover three years and then abandoned. This case illustrates, therefore, the negative policyexperience of a set of Sub-Saharan countries. In Malawi the policy problems were relativelymild, and the government acted fairly early. A great deal of progress was made inimproving the policy framework, and the effort continues. Substantial economic recovery isstill not in sight because (among other factors) the traditional transport route to the sea hasbeen disrupted. Malawi's experience typifies, therefore, the mixed outcome of reform effortsin many African countries. Finally, the Mauritius case is a success story. The government tookearly action to deal with its relatively moderate policy problems. Combined with good luck,these reforms yielded what appears to be sustained economic recovery. There are not manysuccess stories yet, but it is instructive to dissect this relatively rare case.

This book is not for the fastidious reader who insists on formal and rigorous techniques ofanalysis. It is addressed to readers who are keenly interested in policymaking and in policyimplementation. In this universe it is necessary to combine analytical insights from a numberof fields with experiential or practical knowledge (Etzioni 1985). The approach ismultidisciplinary (economics, political science, public administration, and the policysciences), and the methodology is eclectic. I have not shied away from topics that arerelevant but for which rigorous methodologies do not exist or cannot be applied because ofthe weakness of the available data.

Despite this breadth of treatment, I have had to limit the scope of the book in a numberof ways. First, I focus on the political determinants of economic policy, defined as actionstaken to achieve certain goals. I could have explored how the economic situation and otherfactors influence political variables. For example, why is it that democracy survived inMauritius, or why did different kinds of authoritarian rule emerge in other samplecountries? But I do not answer these intriguing questions. Second, my analysis of how thepolitical history of Sub-Saharan Africa has affected its economic policy covers only theperiod from political independence to 1986. Many observers believe that colonial history isalso relevant. The colonial legacy left most of Sub-Saharan Africa underdeveloped. Theimperial powers grossly neglected education, agricultural research, or extension programs.Little was done to establish democratic institutions until the very last moment beforepolitical independence. I have not explored this topic in any detail. Important changes haveoccurred in the region since 1986, but these are not addressed except in brief textual orfootnote references. Readers looking for systematic coverage of up-to-date information willhave to turn to other sources.

The book does not advocate any particular view of what economic policies are suitable inSSA. The reader will soon discover my prejudices as policy developments are reviewed insubsequent chapters, but that is incidental. I do not quantify the benefits of policy reformsince those who have attempted such measurement have made very little progress. Themain emphasis is on policy determinants and government processes involved in making and


implementing economic decisions. At issue here are the political and economic risks inherentin reforms (of course, the risks of delaying reforms may be larger) and what can be done tocontain theses dangers in the special context of low-income Sub-Saharan Africa.

The book is written for all those interested in designing economic policies, in decision-making processes concerned with such issues, and in the implementation of economicmeasures. Naturally, there is the risk that in trying to cultivate a diverse group of readerswith varying professional backgrounds and mixed interests I may please no group. Economicpolicy tends to attract controversy because its multifaceted character is difficult to graspthrough a single discipline such as economics or politics. It is easy for all those involved tomisunderstand each other, and they frequently do.

Chapter 2 explores the acuteness of the economic crisis and its causes. The methodologyused is straight from economics. What is relatively new is the crosscountry comparison, notonly of economic and financial indicators for the 1960s and 1970s, but also of exogenous shocks(disturbances over which governments in Sub-Saharan Africa have little or no control) andweaknesses in economic policies. The focus is on macro policies affecting aggregate demand orfactor markets and on sectoral policies affecting agriculture and manufacturing. Allavailable objective indicators are used, but some subjective judgments are unavoidable.Nevertheless, I hope this disaggregated country-by-country exercise helps to resolve someelements of the controversy about the relative role of shocks and policy failures in causingeconomic difficulties.

Chapter 3 discusses determinants of changes in economic policies during the late 1960s andthe 1970s. The focus is on weak economic policies that contributed to the crisis and theirhistorical and political causation. How is policy made, who are the major actors, and whatis the organizational framework within which they interact? The main political trends areexamined since policy is intimately related to questions of power and status. Themethodology in this chapter draws on political science, public administration, and alliedpolicy sciences. It should help economists to gain a better appreciation of the policy universethan that which their own discipline provides.

Chapter 4 analyses comparative policy responses of governments of sample countriesduring the 1980-86 period. Relatively brief sketches are provided, and the emphasis is onthe central thrust of policy, not on the details. The impact of political events and ofconditional policy lending by the IMF and the World Bank is assessed. The methodologyused in this chapter is an amalgam of economics and political science. Chapter 5 is really acontinuation of the same theme except that the analysis draws on three detailed casestudies of Zambia, Malawi, and Mauritius.

Finally, Chapter 6 goes on to derive lessons of recent policy history in Africa. Why wasit that only two out of twelve sample countries managed to secure intensive and sustainedreforms? Is the transformation of authoritarian, personal rule (that is widespread in theregion) a prerequisite for far-reaching economic reforms? What is the role of politicalleaders, and what is the prospect that African leaders will change their approach togoverning? More broadly, what can be done to improve the internal policy process?

Chapter 6 goes on to examine the extent to which weaknesses in the technical design ofreforms exacerbated problems of implementation. In what way were these weaknesses causedby gaps in economic theory or by the lack of reliable data? These questions are answered inrelation to reforms aimed at stabilization, restructuring of the public sector, exchange ratepolicies, agricultural policies, and industrial (including trade) policies. The merits ofcomprehensive and sequential policy packages are also analyzed.

Chapter 6 also looks briefly at the international framework within which economicreforms took place. How well equipped were organizations like the Paris Club, the IMF, andthe World Bank to play their role in supporting these reforms? Did they insist on anexcessively speedy pace of adjustment? Was their time horizon too short? How effective wastheir approach to policy conditionality? What can be done to improve the interactionbetween these organizations and SSA member countries?In this context I want to add a word

In this context I want to add a word of explanation about the World Bank and its

Introduction 5

affiliated agency, the International Development Agency (IDA). After World War II theInternational Bank for Reconstruction and Development (IBRD), also known as the WorldBank, was established. It borrows funds from all over the world on market terms and lendsthem to LDC governments. The International Development Agency, an affiliate of the WorldBank, was established in 1960 to lend on very concessional terms to relatively poor LDCs.The IBRD and IDA have the same staff, management, and Board of Executive Directors. Inthis book I use the terms World Bank and IDA interchangeably in discussing the work of thestaff on the economic policies of SSA countries. I have tried, however, to distinguishbetween IBRD and IDA loans.

The treatment in this book is far from comprehensive. Much more detailed work can beundertaken on the political and administrative determinants of policymaking in eachcountry and on the distinctive features of the policy process relevant to different types ofeconomic policy. Such research in the context of the United States (and many othercountries) has proved to be rewarding from the standpoint of policy development. It is myhope that similar work will be done on SSA countries.

2ECONOMIC POLICIES AND THEIR CONSEQUENCES

The aim of this chapter is to establish the extent to which policy-induced distortionsadversely affected the economies of sample countries. The route toward this objective isindirect, however, and the end of the journey is a qualitative judgment rather than a preciseestimate. The first part of the chapter assesses the seriousness of the economic and financialcrisis in each country. Next an attempt is made to find out the extent to which thedifficulties can be traced to factors beyond the immediate control of the governments of thesecountries. The impact of exogenous shocks through changes in the terms of trade, in theinflow of foreign resources, and in the weather is assessed. Such disturbances have affectedcountries in the sample very differently. Finally, the chapter evaluates the economic policyframework of sample countries. Where these policies are weak, it is likely that they havecontributed substantially to the economic crisis.

The indirect approach in this chapter to determine the relative contribution of shocksand policy failures can be contrasted with the tidy econometric analysis of David Wheeler(1984). His eight-variable regression model accounts for about 90 percent of variation in thegross domestic product (GDP) of twenty-five African countries from 1970 to 1980. Fourenvironmental variables (violence, share of nonoil minerals in exports in 1970, terms of trade,and export diversification) exercised the dominant impact on GDP growth, but four policyvariables (real effective exchange rate, import allocation regime for capital goods, importallocation regime for consumer goods, and ability to preserve balance in trade accounts) alsohad an important effect. At the individual country level, Wheeler drew two conclusions.First, Kenya's relatively high economic growth performance was largely the result of a goodenvironment, while good policy played a dominant role in explaining Malawi's respectableGDP growth. Second, the very low GDP growth rates in Uganda and Zaire were thecombined result of bad policies and a bad environment.

Wheeler's conclusions are interesting, but he is the first to make a number ofqualifications. First, all his policy variables are related to the balance of payments. He wasnot able to include other policy variables relating to agriculture, industry, public finance,and parastatals. Second, as he points out, grave and widespread data difficulties raiseweighty issues in interpreting the results. It was these problems that led me to adopt theindirect (and less precise) procedure in evaluating the impact of policies on the economiccrisis.

The Main Facts

The purpose of this section is to classify countries according to the extent to which theywere affected by the crisis. Much of the information presented in the tables in this chapteris hard to digest, and it is only included in order to summarize the statistical record forthose who wish to refer to it at any time. The reader should focus on Table 2.6.

Tables 2.1 to 2.3 present data in constant prices on growth in GDP, agriculture,manufacturing, gross domestic expenditure (GDE), investment, consumption, exports, andimports. Between 1967 and 1973, there were striking intercountry differences in rates ofeconomic growth of output or GDP. Six countries achieved growth rates of GDP exceeding 4percent per year, while three scored less than 3 percent. The remaining three cases were inbetween these categories. Four countries showed virtual stagnation in the volume of theirexports, while others succeeded in securing substantial expansion. Differences among countries

7


in the pace of capital formation and changes in consumption were even more remarkable. Twocountries showed a decline in the volume of investment, while eight managed increasesexceeding 5 percent per year. I have not tried to explain these differentials here. Suchresearch would be worthwhile, however.

Table 2.1 GDP and Agricultural and Manufacturing Value Added, 1967-73 and 1973-79(average annual percentage growth; 1980 prices)

Agriculture ManufacturingCountry GDP value added value added

1967-73 1973-79 1967-73 1973-79 1967-73 1973-79

Zimbabwe 10.6 -0.9 5 .1a -0.4 1 0 .8 a -1A

Kenya 9.3 4.8 5.9 4.1 10.1 1.2Malawi 5.6 6.3 4.1 6.1 n.a. 5.5Tanzania 5.2 2.5 2.6 -0.1 7.9 3.4Ethiopia 4.3 1.8 2.1 0.1 10.1 1.2Zaire 4.2 -2.5 -1.7b 0.7 3.4b -5.9Mauritius 3.6 7.0 n.a. -3.7 n.a. 6.2Uganda 3.9 _3.9 d 3 .6c -2.3 d 4 .0c -12.4d

Somalia 3.0 5.0 n.a. n.a. n.a. 8.4Madagascar 2.8 0.8 n.a. -0.9 n.a. n.a.Zambia 2.5 0.3 2.4 2.5 8.0 0.5Sudan 1.5 8.7 4.0 n.a. n.a. 7.7

n.a. not availableNote: The underlying data were derived from national accounts in national currencies. Least squaregrowth rates were calculated. The countries are listed in descending order of their average GDP growthrates from 1967 to 1973.

a. 1969-73.b. 1968-73.c. 1965-73.d. 1973-80.

Source: World Bank data.

Unlike the immediate postindependence period from 1967 to 1973, the 1973-79 period wasone of significant slowing down in the pace of economic activity. GDP slowed down in eightcases, and there was corresponding deceleration in many of the other indices. GDP fell inabsolute terms in three cases, and per capita income declined in seven. Similarly, the volumeof investment diminished in five countries and stagnated in two others during this period.Sudan, Somalia, and Mauritius experienced boom conditions with a remarkable pick-up inGDP but an even more striking acceleration in consumption. The sustainability of this boomwas, of course, questionable.

Financial distress accompanied the economic setback (see Tables 2.4 and 2.5). These tablesshow a variety of indicators for different periods. Data are average for 1973 to 1979. Thepeak ratio during this period (or the trough for external reserves) and the ratio for thesingle year 1979 are also shown. Budget deficits, covered by government borrowing fromcentral and commercial banks, expanded rapidly. Peak ratios of such deficits exceeded 5

Economic Policies and Their Consequences 9

Table 2.2 Gross Domestic Expenditure, Total Investment, and Consumption,1967-73 and 1973-79(average annual percentage growth; 1980 prices)

Country Gross domestic expenditure Total investment Consumption1967-73 1973-79 1967-73 1973-79 1967-73 1973-79

Kenya 9.3 5.2 12.5 2.5 8.9 5.0

Zimbabwe 7.3 1.4 5.2 -16.1 7.6 0.2

Malawi 6.6 5.2 17.6 5.5 6.4 5.7

Tanzania 5.2 3.2 8.3 5.7 5.8 3.5

Zaire 5.2 3.0 15.3 0.1 4.6 5.7

Ethiopia 3.9 2.9 0.5 -2.7 3.9 2.8

Madagascar 3.0 0.5 1.5 0.8 3.5 0.7

Uganda 3.8 a 3.7b 2.1a g.8b 4.0a 3.1b

Somalia 2.6 7.2 5.1 -4.2 2.9 7.6

Mauritius 2.2 9.6 9.6 3.7 3.2 9.7

Zambia 0.4 -5.4 5.6 -20.2 1.5 6.0

Sudan 0.3 10.2 -2.3 8.1 -0.2 11.0

Note: The underlying data were derived from national accounts in national currencies. Leastsquare growth rates were calculated. The countries are listed in descending order of theiraverage GDE growth rates from 1967 to 1973.

a. 1965-73.b. 1973-80.


percent of GDP in six cases; these ratios were a multiple of average levels during 1973 to1979. Governments found it very difficult to restrain expenditures in the face of adverseexternal shocks and economic decline. Very few countries during good times had built upreserves that could be drawn down when the tide turned (Gulhati and Atukorala 1985).Macroeconomic management was weak, and scarcely any country pursued a strongcontracyclical policy.

High budget deficits sparked high deficits on the current account of the balance ofpayments as well as high rates of inflation. Peak balance of payment deficits exceeded 10percent of GDP in eight cases-again a multiple of average levels during 1973 to 1979. Theselarge deficits were financed to a substantial extent by foreign borrowing on hard terms and adrawing down of foreign reserves. Peak debt service ratios exceeded 20 percent in two cases,but several countries found it necessary to have their debt rescheduled. The ratio of foreignexchange reserves to imports in 1979 fell to 10 percent or lower in five cases. Double digitinflation became a characteristic feature of many of the SSA economies.

Table 2.6 classifies sample countries into three categories depending on the magnitude ofthe economic and financial crisis affecting them. Four countries were "most severely affected"


Table 2.3 Exports and Imports, 1967-73 and 1973-79(average annual percentage growth; 1980 prices)

Country Exports Imports1967-73 1973-79 1967-73 1973-79

Zimbabwe 10.1 1.4b 6.0 _11.ObZaire 7.8 -2.6 14.0 -13.3Somalia 5.0 3.8 9.0 4.3Ethiopia 4.7 -2.8 0.0 2.0Mauritius 4.7 4.0 7.0 5.0Madagascar 4.6 -1.5 0.0 0.6Kenya 3.4 32 6.0 0.6Sudan 2.7 -4.8 5.0 3.8Malawi 0.8 4.0 8.0 3.4Zambia 0.7 -0.6 -1.0 -6.5Uganda O.oa -13 7C -4.0 ° 5 .1cTanzania 0.6 -4.7b 8.0 1.0 b

Note: Least square growth rates were calculated for 1967 to 1973. For 1973 to 1979, compound growth rates werebased on three-year averages for both the initial period and the terminal period. The countries are listed indescending order of export growth from 1967 to 1973.a. 1965-73.b. 1973-78.c. 1973-80.Source: UNCTAD (various years).

Table 2.4 Indices of External Financial Disequilibrium, 1973-79

External deficita Debt service ratiob External reservesCCountry Average Peak 1979 Average Peak 1979 Average Trough 1979

Malawi 12 26(1979) 26 11 17(1979) 17 29 13(1976) 18Zambia 7 29(1975) S 24 47(1973) 21 14 8(1978) 10Tanzania 7 13(1974) 8 6 10(1979) 10 16 6(1979) 6Somalia 7 16(1979) 16 4 5(1977) 4 41 18(1979) 18Kenya 7 12(1978) 8 6 10(1979) 10 29 18(1975) 38Zaire 6 16(1977) S 9 15(1975) 7d 21 6(1975) 37Sudan 4 10(1975) 6 14 20(1975) 13e 7 2(1977) 6Madagascar 4 15(1979) 15 4 7(1979) 7 16 1(1979) 1Mauritius 4 12(1979) 12 2 4(1979) 4 27 5(1979) 5Zimbabwe 1 3(1979) 3 If 1(1979) 1 18 10(1975) 33Ethiopia 1 3(1977) 2 7 8(1975) 6 70 33(1979) 33Uganda Gg n.a. n.a. 1 1(1973) n.a. 17 8(1974) 12

n.a. not available.

Note: Countries are listed in descending order of average external deficit from 1973 to 1979. Peak or trough years areshown in parentheses.

a. Deficit on current account of the balance of payments excluding official transfers as percentage of GDP. S =surplus.b. Actual debt service payments as percentage of exports of goods and services.c. Total reserves (gold valued at SDR 35 per ounce) as percentage of imports of goods and services.d. Debt was rescheduled by Paris Club in June 1976, July 1977, December 1977, and December 1979.e. Debt was rescheduled by Paris Club in November 1979.f. Average 1975-79.g. Resource balance was used instead of current account balance.



Table 2.5 Indices of Internal Financial Disequilibrium, 1973-79

Budget deficita InflationbCountry Average Peak 1979 Average Peak 1979

Uganda 3 5(1974) 2 88C n.a. n.a.Zaire 9 18(1976) 4 54 109(1979) 109Sudan 3 6(1977) 5 19 31(1979) 31Mauritius 4 7(1978) 6 15 29(1974) 15Somalia n.a. 12(1978) 8 15 24(1979) 24Ethiopia 2 4(1978) 1 14 29(1976) 16Kenya n.a. 2(1979) 2d 14 19(1975) 8Tanzania n.a. n.a. n.a. 14 27(1975) 14Zambia 5 13(1978) n.a. 13 20(1977) 10Zimbabwe 3 51976) 3f 10 18(1979) 18Malawi 2 3(1979) 3 9 16(1974) 11Madagascar 3e 7(1979) 7 9 22(1974) 14

n.a. not available.

Note: Countries are listed in descending order of the average rate of inflation from 1973 to1979. Peak years are shown in parentheses.

a. Government net borrowing from domestic banks as percentage of GDP.b. Average annual increase in cost of living index.c. 1977-79.d. On a gross basis.e. 1975-79.f. 1976-79.Source: World Bank data.

in that per capita GDP had declined considerably and financial imbalances were large.Uganda had a low external deficit only because it was denied access to external capital, andits foreign exchange reserves had already been drawn down. Zambia and Zaire foundthemselves in a situation similar to Uganda's at the end of the decade, but both hadsubstantial deficits on the balance of payments earlier on. Madagascar experienced anappreciable decline in real per capita income, but its financial situation was not stressfuluntil the very end of the decade. It borrowed heavily on hard terms starting in 1978 as partof the government's policy of "all-out-investment." The debt service ratio in 1979 remainedlow only because it took some time for these hard loans to disburse. Five countries were"substantially affected." In Zimbabwe, Tanzania, and Ethiopia the rate of growth of GDPhad slackened considerably in the late 1970s even though per capita income at the end ofthe decade was higher than at the finish of the previous decade. In Sudan and Somaliagrowth in GDP had accelerated, but both countries had experienced high inflation. Sudan'sexternal debt had become a big burden. Finally, three countries were only "moderatelyaffected" by the crisis. No deceleration in economic growth had taken place in Malawi orMauritius. In fact, GDP had increased at a faster pace from 1973 to 1979 than it had in thepreceding period, but both these countries had encountered considerable financial pressures.By contrast, economic growth had slowed considerably in Kenya, and its balance of paymentson current account had deteriorated considerably.


Table 2.6 Extent of Economic and Financial Crisis, 1970s

Percentage change Budget Rate of External Debtin GDP per deficit as inflation deficit as service

Country capitaa % of GDPb (percent) % of GDPb ratiob

Most severely affected

Uganda -35 ML HH LL LLZambia -21 ML MM HL HHZaire -21 HM HH ML HHMadagascar -15 MH LM MH LL

Substantially affected

Sudan 21 MM MR MM MHTanzania 10 -- MM HH LMSomalia 10 -- H MH HH LLEthiopia 10 LL MM LL LLZimbabwe 7 MM MM LL LL

Moderately affected

Malawi 28 MM MM LL LLKenya 31 -L ML HH LMMauritius 50 MH MM MH LL

Note: Countries are classified into three categories: low (L), medium (M), and high (H).Lack of data is shown by -. The first letter refers to the record over 1973 to 1979 and thesecond letter to 1979 alone. Definitions of L, M, and H are as follows:

L M HBudget deficit Less than 3 3-5 Above 5Inflation Less than 10 10-20 Above 20External deficit Less than 4 4-6 Above 6Debt service ratio Less than 10 10-20 Above 20 or debt rescheduled

a. 1969-71 to 1979-81; constant prices.b. For definitions see Tables 2.4 and 2.5.

Source: Tables 2.4 and 2.5 and Gulhati and Yalamanchili (1988, p. 89).

The Role of ShocksSub-Saharan countries are especially exposed to shocks of various kinds. Most are open

economies, and foreign trade is large in relation to GDP. Fluctuations in international pricesgenerate shock waves in SSA economies. In addition, irregularities in the inflow of officialdevelopment assistance (ODA) and in net resource transfers from foreign loans onconventional terms cause complications. Superimposed on these disturbances are variations inweather conditions that take their toll on agricultural production since SSA countries relyheavily on rain-fed agriculture. The amount and timing of precipitation tend to be irregular.Furthermore, armed conflict and civil unrest have been frequent occurrences in many places.

Table 2.7 summarizes the results of four studies that assessed the magnitude of exogenousshocks. These findings are inconsistent to some extent owing to differences in the authors'approaches and to lack of uniformity in the periods studied.


Table 2.7 Magnitude of Exogenous Shocks

Terms of trade Gulhati andCountry Wheelera indexb Balassac Yalamanchilid

Large impact (L)

Zambia 9 12 (-47) 7 7Uganda 8 2 (7) e 11Zaire 7 11 (-38) 5 9Madagascar 6 7 (-13) 2 5

Medium impact (M)

Sudan 5 4 (-1) 3 6Somalia 4 10 (-26) e 1

Malawi 3 8 (-21) 9 10Ethiopia e 6 (-3) 1 4Zimbabwe e 9(-25) e 3Mauritius e 5 (-2) 8 12

Small impact (S)

Tanzania 2 1 (17) 6 2Kenya 1 3 (-1) 4 8

Note: The higher the rank number, the larger the magnitude of the shock. Countries are listed indescending order of shocks as assessed by Wheeler (1984).

a. Assessment took account of changes in barter terms of trade, coups, share of nonoil minerals inexports, and extent of diversification of exports. The period covered is 1970 to 1980.

b. The rank number is based on percentage change in terms of trade between 1970-72 and 1978-80 asestimated by UNCTAD (1982). Percentage change in the terms of trade is shown in brackets.

c. Assessment took account of changes in terms of trade and slowdown of foreign demand for exports.The period covered is 1971-73 to 1974-78.

d. Assessment took account of changes in terms of trade, net official development assistance, netresource transfer on account of loans from International Bank for Reconstruction and Development,supplier credits and foreign commercial banks, and crop production. The period covered is 1977-79 to1979-81.

e. Not covered by study.

Sources: Wheeler (1984); Balassa (1984); UNCTAD (1982); and Gulhati and Yalamanchili (1988).

Four countries were classified in the "large impact" category. Zambia and Zaire suffereda major deterioration in their terms of trade. Madagascar's terms of trade also encountered aconsiderable setback. Uganda experienced political instability and a great deal of violence.

Six countries were classified in the "medium impact" category. Somalia, Zimbabwe, andMalawi were exposed to substantial terms of trade losses. There was armed conflict in


Somalia, Zimbabwe, Ethiopia, and the Sudan. Mauritius was victimized by cyclones anddroughts. Its terms of trade improved considerably from 1973 to 1975 and then declinedsharply. In the "small impact" category Tanzania and Kenya suffered from rainfall failuresand from the dislocation caused by the breakup of the East African Common Market. Inaddition, Tanzania had to bear the burden of Idi Amin's incursion into its territory and thesubsequent war in Uganda. Although Tanzania and Kenya were subjected to adverse shocks ofrelatively low intensity, the economic impact of these shocks was far from trivial.

The Role of Policy FactorsMany government interventions adversely affect the textbook objectives of economic

policy-economic growth, poverty alleviation, or maintenance of financial balance. Theseadverse effects can be immediate or gradual. Some policies, such as raising investment, canbe good for economic growth, but they may be bad for financial balance under certaincircumstances. A small measure of financial imbalance may do little harm in the short run,but if it persists, and particularly if it grows in magnitude, it can damage economic growth.This simple example illustrates that the evaluation of a single policy is a complicatedmatter even if the social preference function is undimensional. It is much more difficult, ofcourse, to assess an entire policy framework, particularly if there are multiple objectives andtrade-offs that are not clearly specified. Despite these formidable conceptual difficulties, itis still possible to convey some picture of policy problems that had accumulated in thesample countries during the 1970s.

Statist Development

The state attempted to play a leading role in the economies of many sample countries.Unfortunately, these efforts left a great deal to be desired in terms of economic efficiencyand financial management. Government expenditure as a share of GDP rose very considerablyin eight countries (see Table 2.8). The volume of government investment increased at a rateexceeding 10 percent per annum in six countries, which led to budgeting pressures in most ofthem. Many projects were not well prepared or screened on the basis of economic criteria.There was considerable evidence that investment was deployed wastefully (World Bank1984, p. 24; Ghosh 1986, p. 470)1 Economy-wide incremental capital-output ratios roseappreciably in Zaire, Uganda, Zimbabwe, and Zambia (Gulhati and Datta 1983). Partly,this was the result of very limited institutional capacity at the beginning of theindependence period. According to Van Arkadie and Durufle (1986, p. 21), "planning teamswere quite small, involving a handful of staff in central planning agency, minimal staff insectoral ministries and little or no capacity in project analysis or planning at the regionallevel." Over time there was a considerable expansion of the local staff who had acquired afirst degree in economics and who had been trained abroad for a year. But the authorsconclude that "despite more than two decades of experience, the planning function has notyet been effectively incorporated and consolidated as an effective part of the normalmachinery of government' (p. 29). Instead, the planning activity has been marginalized, andplans have become little more than a catalogue of ad hoc projects.

As government expenditure increased, so did the number of jobs in the public sector. By theend of the 1970s, public sector workers as a percentage of nonagricultural employees reachedhigh levels in some countries (see Table 2.8). The public sector became the employer of lastresort in a number of sample countries. A good example was Somalia, which introduced in1972 its guaranteed employment policy. Under this policy, the government undertook toabsorb all secondary school leavers not employed by the private sector. Many SSA

1. Ghosh (1986) provides a detailed account of low-yielding investment projects financed by foreigndebt in Zaire, including the notorious Inga-Shaba complex.

Table 2.8 Characteristics of the Public Sector, 1970s

Category Uganda Zambia Zaire Madagascar Sudan Tanzania Somalia Ethiopia Zimbabwe Malawi Kenya Mauritius

Change in governmentexpenditure as percentageof GDP during 1970sa -16 -2 -11 8 2 17 24 10 8 10 6 11

Percentage change in

government investmentduring 1970s (constant1970 prices)a -23 11 -8 36 13 18 14 -2 -15 21 6 24

Number of parastatals(end 1970s)b 130 114 138 136 138 400 44 180 n.a. 101 176 24

Public sector employees as

percentage share ofnonagricultural employees

(end 19709)C 42 81 n.a. n.a. n.a. 78 n.a. n.a. 20t 39 39 53

Change in public sector

employment during 1970s(percent per year)c n.a. le 15 7t 3e n.a. 12e 7e n.a. 8 6d 8t

n.a. not available.

a. See Gulhati and Yalaminchili (1988).

b. Nellis (1986).c. Growth in public employment. See Lindauer (1988); Heller and Tait (1983).

d. General government only. Excludes parastatals.

e. Central goverrnment only.


governments hired too many workers after the economic setback depressed employment in theprivate sector. Rapid expansion of jobs complicated government wage policy, particularlywhen budgets came under pressure. Many governments opted to retain workers, even thosewho were redundant, and they allowed real wage levels to fall. The burden of adjustment inwages for government workers was larger than for the national average (Lindauer 1988, p.13). A sharp compression in the government wage structure occurred since cost of livingincreases for high-paid workers were offset to a considerably smaller extent than for low-paid workers. Government capacity to recruit professionals tended to decline, and vacancyrates at higher grades increased. Furthermore, recurrent nonwage outlays were squeezed,making it difficult for government workers to carry out their functions. Agricultural extensionworkers, for example, could not travel because of the scarcity of funds, and teachers could notteach because of the shortage of textbooks. "Ghost teachers" in Zaire received salaries eventhough they did not show up for work.2 In Uganda and some other countries, government realwages fell to such a low level that "moonlighting" became a routine.3

Parastatals multiplied rapidly in many countries. Table 2.8 shows the number existing atthe end of the 1970s. At the end of the decade, they were generally responsible for providingutility services, most of which were natural monopolies. Nationalization of foreign miningoperations led to the establishment of parastatals in Zambia and Zaire but not in Botswanaand Zimbabwe, which also have large foreign-owned mines. Agricultural marketing becamea parastatal function during the colonial period, and this arrangement was continued.Farming remained in private hands almost everywhere, however. Parastatals acquired asubstantial part of the manufacturing sector in Ethiopia, Madagascar, Tanzania, andZambia. Altogether, there were more than one hundred parastatals in each of nine countriesout of twelve in the sample.

Examples of well-run parastatals exist; often cited as successful enterprises are EthiopianAirlines, the Kenyan Development Tea Authority, and the Botswana Meat Commission. Thegeneral experience, however, was very negative. Parastatals imposed a heavy financial andeconomic burden on many economies. In Tanzania more than one hundred parastatals wererunning at a financial loss in the late 1970s. In Kenya US$1.4 billion was invested inparastatals, and the average return was only 0.2 percent per annum. In other countriesaccumulated losses of parastatals rose to a substantial level.

Many of the problems affecting parastatals were not of their own making. They were thebyproduct of economy-wide government policies that victimized parastatals as well asprivate enterprises. In addition, parastatals had to contend with interference by governmentin their day-to-day management. Frequently there was a lack of clarity in parastatalmandates. The instructions given by controlling authorities conflicted, and there was lack ofautonomy or accountability. Parastatals faced a "soft budget constraint" (their losses werefinanced by government subsidies or in other ways). They did not have incentives to stand ontheir own feet. Furthermore, parastatals suffered from internal weaknesses such as shortagesof skilled personnel, particularly accountants and financial analysts. Employees tended tofunction as civil servants with a focus on form and adherence to rules rather than on businessoutcomes. Typically, turnover rates at middle and upper management levels were high, andthere were many vacancies.

At the time of independence it was expected that the state would play a criticalentrepreneurial role in capital accumulation and technological progress. These expectationswere frustrated to a large extent. Difficulties in running the newborn parastatals wereattributed to "teething" problems that were expected to be resolved as experienceaccumulated. Policymakers did not, therefore, carry out any midcourse corrections during the1970s.

2 Ghosh (1986, p. 474) estimates the proportion of fictitious names at 65 percent to 70 percent.a 'The civil servant had either to survive by lowering his standards of ethics, performance anddutifulness or remain upright and perish. He chose to survive" (Uganda, Republic of 1982).


Passive Exchange Rate Policy

For much of the 1970s, sample countries tied the value of their currencies to a majorconvertible currency. After the Bretton Woods system of fixed exchange rates broke down,these convertible currencies started to float against each other. The price of foreign exchangein sample countries was a result of these exchange rate fluctuations in the U.S. dollar, theBritish pound, or the SDR (Gulhati, Bose, and Atukorala 1986). Such a passive policy mighthave worked reasonably well if the pace of inflation in sample countries was roughly thesame as in the convertible currency countries. This was not the case, however. During theperiod after the first oil price increase in 1973, inflation in industrialized countriesescalated, but the pace of price increases was even faster in some sample countries. The realexchange rate appreciated considerably in Zaire, Somalia, and Sudan. The rate lost alltouch with reality in Uganda. Even in other sample countries, the official rate did notreflect the growing scarcity of foreign exchange. Declining terms of trade and theaccumulation of debt had exacerbated balance of payment pressures. Several governmentsstarted to ration foreign exchange. While no country maintained payment restrictions onimports in 1967, almost all had adopted such practices by the end of the 1970s.4

The administrative allocation of foreign exchange imposed heavy costs. It was not easy todefine well-specified allocation criteria or to collect timely and reliable data with which toadminister these criteria. Predicting the availability of foreign exchange proved to be amajor problem. Scarcity of qualified professional staff to manage the rationing process addedto the difficulties. Since foreign exchange commanded a large premium in the black market,firms and government officials tried to corner as much of this rent as they could.5 Economicagents spent much more energy in securing import licenses from government and in playing theblack market than in improving productivity or in cultivating new markets for their goods.Applicants for licenses had to complete a lot of paperwork and contend with uncertainty aswell as delay. Many reacted by overinvoicing or underinvoicing, bribing officials, or going tothe black market. Smuggling of traded goods caused governments to lose taxes, and controlover foreign exchange weakened.

A passive exchange rate policy meant that the economic system failed to generatepowerful incentives for earning or saving foreign exchange. Producers of export crops obtainedprices in home currency that did not keep pace with prices they had to pay for importedinputs or consumer goods. Such a deterioration of the local terms of trade of exportersundermined their incentives. The shares of several sample countries in the world market formany of their traditional exports declined. There was little encouragement to identify anddevelop new export items.

Overvalued exchange rates also led to low local currency prices of imports (except forprotected items) compared with prices of competing home goods and nontradables. Incentivesshifted toward the use of capital (since machinery was largely imported) rather than labor(which is a nontradable item). Correspondingly, imported intermediate goods became moreprofitable to use than home-produced materials (Gulhati and Sekhar 1981). Furthermore,consumption of imported cereals, such as wheat and rice, displaced locally produced maize,cassava, and sorghum.

46 For a historical review of the incidence and severity of quantitative restrictions on imports and onpayment restrictions, see Gulhati, Bose, and Atukorala (1985, Appendix 2, p. 51).5. Wood (1988) has calculated the average differential between the official and the black marketexchange rate from 1973 to 1979 as follows: Zaire 75 (the official rate was 25 percent of black marketrate), Zambia 67, Tanzania 51, Zimbabwe 47, Sudan 43, Uganda 40, Malawi 39, Ethiopia 37, and Kenya15.


Discrimination Against Agriculture

Agricultural production stagnated or declined during the 1973-79 period in at least sevensample countries (see Table 2.1). This record was much worse than what had happened from1967 to 1973. Equally striking were the very rapid increases in farm output in Kenya andMalawi during both periods.

Economists could not agree on the reasons for agricultural growth differentials in theAfrican context. They differed in the emphasis placed on price policies, technologicalvariables, and physical and human infrastructure. Uma Lele (1988) explains Kenya'ssuperior past performance in terms of a better initial resource endowment and a number offavorable price and nonprice policies. The future prospect of Kenyan agriculture, however,was clouded by explosive population growth and the exhaustion of high potential land. Asuitable technical package for cultivating semi-arid areas (which have relatively limitedpotential) had not emerged by the end of the 1970s. As I will explain in Chapter 5,Malawi's development strategy brought about a rapid expansion in estate agriculture, butthis was at the expense of smallholders to some considerable extent.

Table 2.9 Nominal Protection of Taxation of Agriculture, 1976-80 and 1978-80

Selected export crops All cropsbCountry 1976-80a 1978-80b 1978-80

Malawi 0.74 0.51 0.49

Tanzania 0.59 0.59 0.64

Ethiopia n.a. 0.46 0.76

Zambia 0.79 1.06 0.84

Kenya 0.97c 1.13 0.93

n.a. not available.

Note: Nominal protection is defined as the price paid to producer as a proportion of border price parity(that is, exports f.o.b. or imports c.i.f. depending on whether crop is an export or import substituterespectively) minus cost of transport, marketing, and processing. A coefficient of less than one indicatestaxation and of more than one subsidy.

a. Unweighted average derived from World Bank (1981, p. 56).

b. Weighted average derived from Kerr (1985).

c. 1971-75.

In the rest of the sample countries, development strategies tended to emphasize rapidindustrialization and to neglect or penalize agriculture. 6 Frequently, producer prices werecontrolled at relatively low levels in order to extract export taxes or marketing boardsurpluses (Eicher and Baker 1982, p. 53). Table 2.9 summarizes available estimates of the

6. Zimbabwe was something of an exception. Economic policies were not biased against agriculture,and the large commercial farming sector achieved high productivity. Until independence, governmentpolicies had neglected the Tribal Trust Lands, however, thereby giving agriculture a sharply dualisticstructure.


extent to which producers were taxed in five countries. Public revenues extracted fromfarmers were seldom ploughed back into agriculture. On the contrary, budgetary outlays oninfrastructure investment or current services for the agriculture sector tended to be inadequate.These policies led farmers to shift from the production of crops whose prices were controlledto other activities. The frequent response to an overall deterioration in the internal terms oftrade of farmers was migration to urban areas.7

Marketing parastatals were assigned a dominant role. In many cases these stateenterprises enjoyed a near-monopoly legal status. The case for such a policy was that itwould allow government to subsidize production and trade in remote areas as well as tostabilize prices and to ensure the availability of food throughout the country. Activities ofprivate traders were circumscribed by a variety of government restrictions, includingtransport barriers impeding interdistrict movement of produce. Marketing parastatals tendedto have relatively high administrative and operating costs (Eicher and Baker 1982, p. 55).Furthermore, they could not stabilize prices and provide food security because their legalmonopoly power was eroded by the emergence of illegal operators. As the capacity ofparastatals to purchase from farmers diminished, their costs per unit increased even moresince heavy overhead costs had to be covered by a reduced volume of transactions.Furthermore, illegal transactions were risky and tended to discourage long-term investment ininfrastructure for private marketing.

Many countries subsidized agricultural inputs (particularly fertilizer) and distributedthem through parastatals. The aim was to promote the use of such inputs, which wereintroduced for the first time in many instances, and to offset the impact of low producerprices. Benefits of input subsidies accrued mainly to commercial farmers and to therelatively affluent small farmers. The great majority of traditional farmers seldom obtainedsuch subsidies. Consumer prices for major cereals were also subsidized in several instances.These subsidies became a major burden on the budget and posed severe policy problems, asillustrated in Chapter 5 for Zambia, Malawi, and Mauritius. Benefits of subsidized priceswere available to the poor and the affluent. Actual availability of staples in ration shopsvaried from place to place and from time to time. A sizable portion of consumption of grainsdepended on unofficial operators selling at prices much higher than the official subsidizedlevel. Correspondingly, there was a shortfall in building and maintaining feeder roads thatwere critical for the commercialization of agriculture. Considerable amounts were spent onagricultural extension, but Lele (1987) argues that there was little pay-off except in the caseof hybrid maize in Kenya. She also maintains that the productivity of outlays onagricultural research was low (except in Kenya and Malawi in the case of export crops)because of poor research policies.8

This stylized policy regime was hardly peculiar to Sub-Saharan Africa. A study byKrueger, Schiff, and Valdes (1988) covering eighteen countries drawn from all less developedcountry (LDC) regions confirms that agriculture was the victim of discrimination almosteverywhere. On average from 1975 to 1979, agricultural exports were taxed at a rate of 11percent, while food items were protected at a rate of 20 percent. The impact of these direct

7. For example, domestic terms of trade of export crops in Kenya declined by 41 percent from 1970 to1979 (World Bank 1981, p. 174). During this period, Kenya's urban population increased at an annualrate of 6.8 percent compared with 3.4 percent nationwide.& "The problems include the lack of adequate training and experience of African scientific staff, bothin research and management, excessive reliance on expatriate staff (exceeding 50 percent of thenational scientific establishment in several countries), and frequent changes in the organization ofresearch systems. Organizational changes are often driven by domestic political imperatives.Conflicting advice from a fragmented donor community that embodies quite different approaches toresearch exacerbates the problem. Little relationship exists, therefore, between the organization, themanagement, and the substance of national research systems. The expatriate experts provided bydonors to alleviate the shortage of trained African manpower tend to turn over quickly, which inhibitsleaming by doing. Also, the manpower provided by donors is geared more to doing research themselvesthan to training Africans to do research" (Lele 1986, pp. 31-32).


policies was much less, however, than the impact of indirect interventions such as exchangerate and trade policies affecting manufactures. Taking account of all policies during the sameperiod, agricultural exports were taxed on average at a rate of 36 percent and food at 5percent.

High-Cost Industrialization

Very few sample countries had a static comparative advantage in manufacturing whenthey became independent. Domestic markets were relatively small owing to smallpopulations and low per capita incomes. There was an acute shortage of engineers, managers,and skilled workers except in foreign-owned companies. Despite these unfavorable initialconditions, most countries adopted ambitious industrialization objectives. The industrialsector was seen as a leader in the development strategy by African planners. It was to bringabout modernization and technological progress. The aim was to replace importedmanufactures by local production. A second aim in some sample countries was to secure controlover foreign-owned companies by nationalizing them.

A typical policy regime for the manufacturing sector included four major elements. First,extremely high nominal tariff protection was given to locally manufactured consumer goods.Rates were as high as 300 percent in Uganda and 220 percent in Kenya. Correspondingnominal duties on intermediate and capital goods remained at low levels, thereby leading toa wide dispersion in the structure of tariffs. Furthermore, the number of items on theprohibited or restricted list of imports expanded. In most cases tariff duties and quantitativerestrictions were imposed for indefinite time periods. Second, as the balance of paymentcame under pressure, licensing of imports became more and more restrictive. In many instancesthe primary aim of quantitative restrictions was to manage foreign exchange, but an indirectresult was higher protection to local producers of competing goods. Third, as importprotection spread, governments became concerned about local industrialists chargingexorbitant prices; governments then imposed controls on prices of items they considered to beessential. Fourth, the number of parastatals engaged in producing manufactures rose rapidlyin many sample countries as a result of nationalization of foreign firms and the starting ofnew firms in the public sector.

Three of the sample countries-Kenya, Tanzania, and Uganda-were members of the EastAfrican Common Market (EACM) until it collapsed in 1977 (Gulhati and Sekhar 1982). Afree trade arrangement during the colonial period was continued after independence.Although Tanzania had considered withdrawing from this arrangement in the early 1960s,it finally decided to continue, subject to four alterations (Hazelwood 1975, p. 57). First, afiscal scheme to compensate Tanzania and Uganda for net losses accruing to them from thecommon market arrangement was introduced. Second, a number of industries were allocated toTanzania and Uganda on an ad hoc basis, and all three member governments agreed toundertake planning studies leading up to further allocations of new industries. Third, theEast African Development Bank was established not to promote industries in all threeEACM members but to make preferential allocations to relatively backward Tanzania andUganda. Finally, Tanzania and Uganda were allowed to impose transfer taxes on importsfrom Kenya that competed with their own production. Notwithstanding all thesealterations, the East African Common Market did not flourish. During its heyday, however,Kenyan and Tanzanian manufacturing exports to the EACM were a sizable part of production.Then, in 1977, trade between the two countries came to a standstill. An effort was made torevive these arrangements in 1978 when twelve states (including most sample countries)signed a letter of intent to establish a Preferential Trade Area. These arrangements did notcome into operation until 1982, however. To qualify for preferential treatment, the exportingfirm must have majority management and ownership of national origin. The impact of thesenew preferential tariff arrangements was reduced considerably by import licensing andnontariff restrictions.


A regional arrangement has the potential for furthering industrialization in samplecountries, all of whom have very small domestic markets. But regionalism in Sub-SaharanAfrica faces many severe constraints. These economies are not complementary. Theirinfrastructure is geared to trade with erstwhile metropolitan centers rather than with eachother. They are preoccupied with managing their national economies, and they face seriousshortages of professional manpower that would be required to design and administer regionalarrangements. Political capacity for initiating and managing such arrangements is lacking.

By and large, manufacturing catered almost exclusively to the home market at the end ofthe 1970s. Many investments, financed by foreign debt on near-market terms, were undertakenwithout proper technical, financial, or economic screening. A large number of these projectswere capital intensive and import intensive. The capacity of some projects was too big forthe home market, and their high costs ruled out overseas sales. Production for the homemarket depended on the availability of substantial amounts of foreign exchange for purchaseof intermediate goods and spare parts from abroad. But as the economic crisis unfolded,foreign exchange became increasingly difficult to obtain. There occurred a marked fall incapacity utilization toward the end of the decade, further adding to inefficiency andalready high production costs.

Experts familiar with the manufacturing sectors of sample countries tend to classify themin three groups based on performance. Table 2.10 summarizes indicators of efficiency of themanufacturing sector in the twelve sample countries. Despite well-known methodologicalweaknesses of the indicators, gaps in available information, the varying number of productsused in the study samples, and the different years covered, rough crosscountry comparisonscan be made. The first four countries in the table had relatively efficient manufacturingsectors. The average domestic resource cost (DRC) of earning or saving foreign exchange wasvery near one (on the basis of shadow exchange rates) in Zimbabwe and Kenya. About 10percent of Zimbabwe's manufactures and 8 percent of Kenya's was exported. However, therewas a very large variation around the average, particularly in Kenya. Segments ofmanufacturing in these countries enjoyed high effective rates of protection (EROP) andoperated at rather low levels of efficiency. Malawi and Mauritius also belonged to thisgroup, although Malawi's average DRC was somewhat higher. More than a fifth ofMauritius's manufacturing output was exported. Its Export Processing Zone was a spectacularsuccess in the mid-1970s, but its growth slowed down at the end of the decade owing to largewage hikes (see Chapter 5). The rest of Mauritius's manufacturing sector catered to the homemarket, some of which was heavily protected.

Ethiopia, Somalia, Sudan, and Madagascar faced larger problems of inefficiency. In Table2.10 the average DRC for Ethiopia is a relatively low figure, but it is based on a smallsample and therefore is not very reliable. The average EROP for Madagascar is low, but it isbased on a very old study, and it is well known that protection rose sharply later. Finally,the last set of four countries in the table had even larger problems of inefficiency in themanufacturing sectors. Tanzania, for example, had an average DRC at the shadow exchangerate of 2.91. DRC estimates for Uganda are misleading because they do not take into accountfixed costs of production.

Summary

The sample countries were classified into three groups based on a survey of economic andfinancial indicators (see Tables 2.1 to 2.5). The classification "most severely affected" by theeconomic crisis, "substantially affected," and "moderately affected" is presented in Table 2.6and retained in Table 2.11. Next I examined the findings of a number of studies that assessedthe magnitude of exogenous shocks, and I showed a rank order of countries in Table 2.7. Theclassification in column one of Table 2.11 is based on this earlier analysis. Finally, afterreviewing available indicators of policy problems in four major areas, I arrived at theclassification in columns 2 to 6 in Table 2.11.


Table 2.10 Indicators of Efficiency of Manufacturing

Effective rate of protection

Country and year No. of products Domestic resource costa at official exchange rateb

of study in study sampleOfficial exchange rate Shadow Exchange Rate

Average Range Efficient

productsc Average Average Range

Zimbabwe (1981) 33 1.27 0.68-3.62 16 1.05 33 -14 to 177

Kenya (1985) 106 1.53 0.71-6.29 47 1.08 89 6 to 312

Malawi (1987 100 1.65 0.50-6.20 52 1.48 148 38 to 641

Mauritius (1982) n.a. n.a. n.a. n.a. n.a. 89 -24 to 824

Ethiopia (1983) 19 1.09 0.36-4.37 6 0.84 36 -53 to 120

Somalia (1985) 15 n.a. 0.14-2.99 1 n.a. n.a. -236 to 641

Sudan (1986) n.a. n.a. n.a. n.a. n.a. 240 57 to 690

Madagascar (1 9 76 )d n.a. n.a. n.a. n.a. n.a. 82 34 to 112

Uganda (198 3)e 44 n.a. 0.03-3.05 26 n.a. 178 29 to 821

Zambia (1981) 48 n.a. 0.47-3.02 17 n.a. n.a. n.a.

Zaire (1985) n.a. n.a. n.a. n.a. n.a. 300 -20 to 1,800

Tanzania (19 84 )f 135 n.a. 0.40-11.07 46 2.91 470 53 to 3,780

n.a. not available

Note: Countries are listed roughly in descending order of effidency of their manufacturing sectors.

a. Of earning or saving one unit of foreign exchange. Calculations are based on the official exchange rate and the shadow rate (taking

account of tariffs, subsidies, and the overvaluation of the official rate). The average is weighted by the share of various branches in

total value added of the manufacturing sector.

b. Takes account of differentials between domestic and border prices of outputs and imported inputs. The average is weighted as in (a)

above.

c. Number of products with a DRC of less than one, based on official exchange rate.

d. Protection rose substantially later in the 1970s and in the 1980s.

e. The DRC and EROP are based on offidal rate (window one) after a major devaluation in 1983. The DRC is based on short-run costs

only.

f The DRC and EROP are calculated at shadow exchange rate.


My approach has been to proceed as far as possible on the basis of available data andobjective analyses. These elements are weak, however, and at some point I had to "take aleap" in order to arrive at the conclusions summarized in Table 2.11. These conclusions arecontroversial. Two main points should be emphasized, however. First, sample countriesshared a number of common characteristics. They were exposed to exogenous shocks that, onbalance, had an adverse impact during the 1970s. Their policy framework tended todeteriorate particularly in four areas specified in Table 2.11. Second, the incidence of shocksvaried a great deal among sample countries as did the acuteness of policy-induceddistortions. These conclusions are useful reminders of the value and limitations ofgeneralizations about Sub-Saharan Africa. The range of variation among countries in thisregion is enormous, and yet the negative economic trend characteristic of the whole region isundeniable. The following chapter will address the political determinants of this trend andthe variations around it.


Table 2.11 Economic Profile of Twelve Sample Countries, 1970s

Impact Impactof of policy Public sector Exchange rate Bias against High-cost

Country shocks problems inefficiency overvaluation agriculture industry

Most severely affected by crisis

Uganda L L L L L LZambia L L L L L LZaire L L L L L LMadagascar L L L L L M

Substantially affected by crisis

Sudan M L L L L MSomalia M L L L M MTanzania S L L L L LZimbabwe M M M S M SEthiopia M M M S L M

Moderately affected by crisis

Malawi M S S S M SKenya S M M M S SMauritius M S M S S S

Note: The contribution of various factors to the crisis is assessed by classifying them intothose that had a large (L), medium (M), and small (S) impact.

Source: Based on Tables 2.1 to 2.10.

3IMPACT OF POLITICS ON ECONOMIC POLICY

At the end of the 1970s, many sample countries were burdened not only by unfavorableinitial conditions reflecting their stage of underdevelopment, but also by serious "policydistortions" and the impact of negative exogenous shocks. "Distortions" were defined inChapter 2 as government interventions that impeded progress toward the usual textbookobjectives of economic policy-growth in GDP, equal distribution of income, and financialstability. What happened in these countries (that is, the accumulation of distortions) wasthe opposite of what was suggested in the early postwar period by theories underlyingapplied welfare economics or development economics.

These theories implicitly regard the state as a guardian of the long-term interest ofsociety.1 Government interventions, according to these views, can correct "market failures,"thereby enabling society to achieve its objectives (for example, efficiency, growth, equity) toa larger extent than under a laissez-faire policy.2 Theories of welfare and developmenteconomics appear to assume that the government is monolithic and that it has anunambiguous preference function that coincides with that of society. They make nodistinction between (1) personal objectives of civil servants and politicians, (2) bureaucraticimperatives of government ministries, (3) vested interests of various economic or ethnicgroups, and (4) the aggregate social preference function. Little attention is given to datalimitations and weaknesses of analysis in diagnosing market failures or to problems of publicadministration in implementing economic policies.

Given this framework, which Allison (1971) calls the "rational actor" model, it was noteasy to understand why governments invested wastefully during the 1970s and why theyreduced outlays on transport or textbooks, which were necessary for government extensionworkers and teachers. Equally puzzling was the fact that during this period governmentsgave conflicting instructions to parastatals, maintained overvalued exchange rates, andrationed foreign exchange in the absence of reliable data and skills required for this purpose.More examples of this mismatch between what was expected from governments by welfare ordevelopment economists and governments' actual behavior could be added to this list. Onemust conclude that the analytical framework that leads to such misleading inferences aboutpolicy needs a basic revision.

This chapter is based on an alternative analytical framework drawn from a number offields: economics, political science, theory of public administration, and policy sciences. Thisframework is defined by four propositions. First, the economic policy process is greatlyaffected by political factors such as who is in the policymaking circle and how much effortis made to build consensus. A basic distinction is between democratic and authoritarianpolitical systems. Democracies can be classified as majority party or coalition governments,

1. "The general presumption," according to Killick (1989, p. 14), "was that the state was benign in itsintentions, with the theory of policy centered around the question of how best the state could maximizesocial welfare."2. Market failures (or deviations from conditions of perfect competition) are caused by monopolyelements, external economies, public goods, paradox of thrift, and fallacies of composition. Accordingto neoclassical economic theory, such market failures require optimal government interventions.Economists were, of course, aware of the widespread incidence of nonoptimal governmentinterventions, but these facts were not integrated into mainstream economic theory. There was littlesystematic exploration of reasons why governments intervene in ways economists regard asnonoptimal.

25


parliamentary or presidential systems, and proportional representation or single constituencyelectoral systems. A particularly useful distinction is drawn by Haggard and Kaufman(1989, p. 233) between "plebiscitary" democracies (where political parties are mainlyvehicles for getting the vote and much less for representation of interest groups) and"consultative" democracies (where there are institutionalized channels for interest groups tointeract with the executive). Authoritarian states are classified by the same authors into'"strong" and "weak," a point I will develop later. Most states in Sub-Saharan Africa areauthoritarian, and they are characterized by "personal rule" (Jackson and Rosberg 1982). Thecharacter of these rulers influences the economic policy process substantially.

The second proposition upon which my altemative analytical framework is based is thatpolicy is also influenced by social stratification in terms of class, ethnicity, and regionalloyalties. Rulers have a measure of autonomy in the policy sphere, but how much room formaneuver they have depends on the nature of group interests and the political salience ofthese groups. Third, foreign donors and investors influence economic outcomes through thesizable resources they provide and through their ideas about economic policy. Fourth, thesize and quality of the civil service influence policy decisions and their implementation.

These four propositions are interlinked. For example, a strong authoritarian ruler canundermine the organizational effectiveness of interest groups. These groups are likely to playa bigger role in economic policy in consultative democracies than in plebiscitary democracies.The contribution made by the civil service to economic policy is heavily influenced by thepolitical system and by the character of the leadership as well. A strong ruler can deploythe civil service to carry out his preferred policies, or he can use its technical expertise todecide what policies he should adopt. Under a weak ruler or in a democracy, civil servantscan use their policy or administrative role to further their own interests or those of theirministries. A ruler's strength is also affected by the posture of his foreign patrons. Accordingto Hadjor (1987, p. 5), many African governments rely more on the support they get fromabroad than on local backing. Many other examples of interconnections among these fourpropositions can be cited.

The next section of this chapter begins with a discussion of political trends afterindependence. This is followed by an assessment of how these trends influenced the mannerin which policy was decided. The main policy decisions are summarized, and the impact ofpolitical parameters on these decisions is explained. Next I identify the role of politicalvariables in policy implementation. In each section the discussion is in three stages. First,some broad generalizations at the sub-Saharan level are advanced. Second, key elements ofdiversity within the region are noted. Third, illustrations to fortify the argument are drawnfrom the sample countries. This chapter emphasizes the impact of domestic political factorson economic policy. The role of external actors is analyzed more fully in later chapters.

My analysis strikes a rather critical note about developments in Africa afterindependence. Many of these criticisms, however, also apply to other Third World regionsand indeed to advanced countries, particularly in their early stages of development.

Major Political Trends

The colonial state was an institutional structure serving metropolitan interests. Keypolicy decisions were made by European powers and were implemented through abureaucracy. There was scarcely any participation in matters of state by the local people.Except for clerical and junior posts, the bureaucracy was staffed by expatriates. Rothchildand Curry (1978) call this the "command model."

When independence arrived, colonies were ill equipped politically and administratively.Most states were multi-ethnic, and their societies were segmented. Citizens of the new statesdid not have the bonds that come from a shared language, religion, or tradition, althoughparticipation in the anticolonial struggle stimulated national sentiments to some extent.There was an acute shortage of qualified local people to staff middle and senior posts in thecivil service.

Impact of Politics on Economic Policy 27

Just before independence, European powers instituted in the colonies constitutionalstructures involving bicameralism, federalism, and multiparty systems. The aim was toexpand political participation of the people, but these experiments were frustrated in mostcountries. According to Fieldhouse (1986, p. 56), "the result was that only a very partialmetamorphosis of the colonial state and society had taken place before decolonization; thenew nation state inherited a tradition of autocracy barely tempered by democracy."

There was a strong trend toward authoritarianism during the seventies, parallelingearlier developments in Asia. Participatory government in Africa lacked a basis in colonialhistory. Elements of self-government existed in traditional ethnic communities in theprecolonial period, but it was difficult to transplant this experience to the national levelbecause of the artificial administrative boundaries of the new independent states.3 Accordingto Sandbrook (1986, p. 322), over 50 percent of regimes in Sub-Saharan Africa in the mid-1980s were military or quasi-military, and 33 percent or more had one-party governments orhereditary monarchies.

Like medieval regimes in Europe, the typical contemporary regime in Sub-Saharan Africarevolves around the person of the ruler.4 It is noninstitutional and authoritarian. Complianceis secured mainly by a combination of mercenary means (patronage) and coercion. Patriotismand cultural norms and ideology play a relatively small role. There is a great deal offactional struggle and uncertainty is endemic. Patron-client ties bind leaders and followers.Although purely economic interest groups, such as trade unions and commercial farmers, exist,many of them are not well organized. Frequently, they are coopted or repressed by theregime. Military or quasi-military regimes are not much different from civilian governmentswith respect to the way in which they rule. Many rulers are so weak that they do not evencontrol sizable parts of their territory.

In an earlier century such rulers might not have won diplomatic recognition from manyforeign powers (who typically waited for the internal political competition to concludebefore formalizing interstate relations with the victorious ruler). In the period after WorldWar II, however, successors named by retreating colonial authorities were recognized at once.They did not have to emerge as victors over rival claimants to political authority. Theybecame sovereign by virtue of membership in the Organization of African Unity (OAU) andthe United Nations. They acquired juridical rather than functional statehood (Jackson andRosberg 1986).

The character and style of the ruler are important in determining the capacity ofindividual regimes. Jackson and Rosberg (1982), for example, classify these rulers into fourcategories: princes, who tend to share power and respect understandings reached withsupporting factional heads (most SSA rulers fall into this category); autocrats, who refuse toshare power with others and deploy a bureaucracy to carry out their policy commands and toadminister routine functions; prophets, who are visionary and exercise power as aninstrument for securing a higher goal; and tyrants, who do not accept legal or moral

3. Following are excerpts from Bgoya and Hyden (1987), authors of a report on a seminar organized bythe Dag Hammarskjold Foundation and attended by prominent African academics and others. 'Beforecolonization African societies were not organized in 'states' in the modem sense. Centralization ofpower and bureaucratization...had not developed in Africa" (p. 8). "The basic principle in Africanpolitical life which was ignored by subsequent systems was the socialization of political and economicpower" (p. 9). "The state-society relations which developed out of the slave trade and colonialismengendered two realms in constant conflict: an amoral and ill conceived civic realm on one hand and acircumscribed community-based moral realm on the other" (p. 12). "African leaders (afterindependence) turned democracy into mockery and allowed it to degenerate into personality cults,factors that invariably contributed to the phenomenon of the coup d'etat" (p. 15).4. Jackson and Rosberg (1982) present this view. Nuances apart, the same general insight is shared byCallaghy (1986b), Sandbrook (1986), Hyden (1983), and Bates (1983), among others. The "personal rule"view contrasts with the Marxist view of the state as a creature of international capitalism or theemerging local bourgeoisie. For a survey of the literature on the African state, see Marenin (1987) andStark (1986).


restraints on their power and act arbitrarily. These four categories are useful at thedescriptive level. Beyond that, they have some predictive value. For example, princes andautocrats tend to pursue conservative policies, while prophets are prone to adopt radicaldoctrines. And decisionmaking by autocrats is usually more centralized than by the otherthree types of rulers.

There are three problems, however, with the Jackson and Rosberg typology of rulers.First, it is entirely static; it does not focus on factors that cause princes to become autocrats ortyrants and vice versa. This is a major limitation. Second, the category of prophetic rulersdoes not appear to be on the same logical plane as the other ruler types. A visionary rulercan be a prince or an autocrat depending on whether or not he shares power with others.Third, border-line cases are hard to classify since rulers share power to varying degrees.These limitations notwithstanding, the typology is a useful one.

Zambia is an example of a "princely state," while Malawi is an autocracy (see Chapter5). Tanzania's Julius Nyerere was a "prophetic" ruler. He advocated a return to anegalitarian, rural society that predated the arrival of the Europeans in Africa. He alsoadvocated a welfare state. He believed that such a transformation could occur withoutviolence. He was cast in the role of a charismatic leader and a teacher ("Mwalimu").

In making the Arusha Declaration in 1967, Nyerere parted company with mainstreamcapitalist or Marxian doctrines and began to articulate an original ideology and philosophy.He wanted the party to be the instrument of social change. Tanzania had already become aone-party state, and in 1969 supremacy was transferred from the parliament to the party.Elections were held every five years with voters having a choice of "yes" or "no" forNyerere, the only candidate. In each election he obtained over 90 percent of the vote. Votersalso had a choice between two candidates approved by the party for each parliamentaryseat. The president could appoint fifteen members of parliament; on occasion many cabinetmembers were appointed rather than popularly elected. Parliament played a rather passiverole in decisionmaking. The National Executive Committee of the party tended tomonopolize power. It neutralized the autonomy of the trade unions, and it did not permit theestablishment of new interest groups (Pratt 1979). If policies of the party came into conflictwith the judiciary, the party's policy prevailed.

Idi Amin exemplified tyrannical rule. He ousted Milton Obote in a coup in 1971, issued acharter for a prosperous Uganda, and promised restoration of civilian rule in five years. Inpractice, the Amin regime persecuted many individuals from many parts of society butparticularly the Acholi and Langi people who supported Obote. Amin favored the Nubiansand the Kakwa instead. Mercenary soldiers protecting Amin or his entourage were allowedto confiscate private property, sell government property, and terrorize the citizenry. Civilrights of ordinary people were undermined by arbitrary decrees. According to Jackson andRosberg (1986, p. 222), Uganda descended to a Hobbesian "state of nature."

Most African states fall into one of the four "personal rule" categories. Botswana andMauritius, however, illustrate that personal rule is not universal in the African continent.(The Mauritius story is told in Chapter 5.) A multiparty constitutional system has functionedin Botswana for more than two decades, and an independent judiciary enforces a code ofhuman rights. The Botswana Democratic Party (BDP) won all five elections, but severalopposition parties competed and together they obtained a quarter of the vote. Oppositionparties in Botswana have declined politically, but they are not impotent.5 Seretse Khama,the president throughout the independence period, died in office in 1980. He was succeededby Quett Masire, thereby demonstrating systemic capacity to carry out a constitutionaltransfer of power. There is no wholly satisfactory explanation for the survival ofconstitutionalism in Botswana, but scholars have emphasized the contribution of threefactors. First, the traditional institution of the Kgotla (the village assembly where views

5. According to Picard (1987, p. 173), "the ruling BDP uses the advantages of its incumbency and theadministrative mechanisms of the state to maintain its predominant political position within thecountry."


were expressed freely and consensual decisions made) fostered a democratic culture(Raphaeli, Roumani, and Mackellar 1985). Second, there is a deep desire for peace and aprofound respect for the law (Alverson 1978). Third, the national movement was an amalgamof traditionalism and liberalism. It tended to nurture a reformist, rather than a radical,orientation (Picard 1987).

Policy Circle

Not surprisingly, the economic policy process has reflected the political parameters. Inmost SSA regimes there is little political participation, thereby confining policymakingactivity to a narrow circle. This circle is particularly small in autocratic states, asillustrated by Malawi, and somewhat larger in Zambia (see Chapter 5). It may not bemeaningful to talk about a policy circle under tyrannical regimes such as Amin's in Uganda.Such states are characterized by arbitrary actions that undermine rationality,predictability, and certainty in all spheres of life, including the economic. The very notionof policy is questionable in such a context.

President Nyerere dominated the policy circle in Tanzania, particularly in the sphere ofrural development policies. In other fields, Nyerere did not preempt all initiative. Someadvisors with access to the president (Justinian Rweyemanu for example) played a major rolein policies affecting industry. Members of the National Executive Committee of the partycould also play a decisive policy role on some occasions. Finally, the growth of the publicsector created a new ruling elite of public officials (Shivji 1976). They were a diverse groupin terms of income, status, and organizational allegiance, but for some considerable timenearly all subscribed to Nyerere's ideology of Ujamaa. Some of these senior officials playeda significant policymaking role on certain occasions, and a larger group was criticallyinvolved in policy implementation.

The policy circle is much more institutionalized in Mauritius and Bostwana. Chapter 5spells out the impact of parliamentary elections, and of the rise and fall of coalitions amongparties, on the policymaking style of Mauritius. The policy circle in Botswana consists of thepresident, the cabinet, and the civil service. According to Picard (1987), the bureaucrats area significant interest group and play a disproportionately large policy role. The politicalelite allow the civil servants to play such a role because both groups share the same valuesand have similar economic interests-namely, cattle rearing. There is a strong commitment torational decisionmaking. A great deal of effort is made to consult and to come to anagreement with all affected parties, including the private sector (Raphaeli, Roumany, andMacKellar 1984). Both Botswana and Mauritius can be classified as "consultative"democracies using the terminology of Haggard and Kaufman (1989).

Policy Climate

The policy process in Sub-Saharan Africa also reflects the grossly inadequateinstitutional structure in most countries. They have the usual governmental agencies: the coreeconomic ministries, the sectoral ministries, the parastatals, and local governments. MostSSA countries also have the usual interministerial machinery of coordination, of which thecabinet is the most important. All these institutions exist, but they do not function very well.The underdevelopment of human resources for policy work was part of the historical legacyof colonialism. The stock of indigenous professionals to analyze economic problems, to definepolicy options, and to manage the entire process of policy preparation was very low atindependence. There was heavy reliance on expatriate staff. Outside government, there werehardly any university departments or "think tanks" to do policy work. Statistical systemswere very weak, making it unusually difficult to conduct quantitative policy studies.

Africa's higher education system expanded dramatically during the 1970s and 1980s. Thenumber of university graduates in Sub-Saharan Africa rose from 1,200 in 1960 to 70,600 in


1983. Even today, however, the stock of graduates is about 0.4 percent of the population agedtwenty-four and older, compared with perhaps 6 percent on average in other LDCs (WorldBank 1988a, p. 70). A large number of the graduates studied arts and humanities and cannotcontribute much to professional policy work. Even the relatively high-quality educationalinstitutions have been severely damaged by the ongoing economic crisis. There continues toexist a serious shortage of high-quality economists (particularly macroeconomists), managers,and other professionals required for policy work (Ndulu 1986, p. 92). Africa has received alarge amount of foreign technical assistance, but there is limited (and rapidly declining)tolerance for expatriates directly engaged in policy activity. These foreign experts have notsucceeded in transferring policy analysis and management skills to their Africancounterparts.

Moreover, the existing policy talent is not being utilized fully. Political andadministrative conditions do not permit effective deployment of such skills as are availableto governments. The small circle of policymakers tends to make decisions on the basis ofintuition, ideology, or a process of give and take. There is little appreciation of howtechnical policy analysis can feed into the decisionmaking process. Consequently, the rulingcircle has not generated much demand for policy relevant studies. Think tanks outsidegovernment have been discouraged, and little effort has been made to improve the statisticalsystem. Not only is there a demoralizing lack of demand for the services of policy analystsand statisticians, but they face a forbidding work environment (low pay, bad physicalconditions, etc.).

Not much systematic empirical evidence can be cited in support of the bleak picturepresented above. But in 1984 Montgomery (1986a, 1986b, 1987) carried out a useful study ofbureaucratic behavior in nine countries (members of the Southern African DevelopmentCoordination Conference). This investigation concluded that policy issues occupied only aminor part of the work agenda of permanent secretaries who were at the apex of the civilservice. They did little strategic analysis. Work aimed at adapting their organization'smission to the changing environment was much less prominent than might have beenexpected. Bureaucrats tended to respond defensively when confronted with questions ofpolicy. A substantial part of their work was inward looking.

The policy process was very weak in Sub-Saharan Africa, but the severity of the problemvaried from one country to the next. In Botswana and Malawi, for example, there isrelatively more tolerance for expatriates in policy positions. Physical working conditions orincentives did not deteriorate as much as in many other SSA countries. The relationship ofsenior civil servants to the political authorities differed in Malawi and Botswana, however.In Malawi the relationship was one of the utmost delicacy (see Chapter 5). If senior civilservants disagreed with government policy on some topic, they would hesitate to articulatetheir misgivings for fear of being accused of disloyalty. They were reluctant to be the bearerof "bad news" about the economic situation. In Botswana civil servants enjoyed the confidenceof the politicians. Communication between the two groups was easy in both directions.Zambia (see Chapter 5) offers a sharp contrast to Botswana and Malawi. Its policy processwas flawed by acute shortages of professional expertise, interministerial rivalries, and"presidentialism." Working conditions had deteriorated greatly, and communication betweencivil servants and politicians was poor.

Political Parameters and Policy Content

The turnaround in Africa's economic policy during the independence period from about themid-1960s to the end of the 1970s was a sharp one. In some countries the change amounted to


an entirely new policy framework.6 To some extent the turnaround reflected the departure ofthe imperial powers. For example, the involuntary accumulation of sterling balances in theBritish colonies came to an end. Physical controls on land use that favored the foreign settlerpopulation were terminated. A very low tariff policy on imports from metropolitan centerswas wound up. The turnaround was also driven by the political imperatives of "personalrule," although the ideology of rapid socioeconomic development provided the ostensiblejustification. The central motivation behind economic policy was regime survival. Theemphasis was on buying political support and sustaining it, thereby avoiding the risk ofbeing driven into the political wilderness. To the extent that the same policy could also berationalized in terms of the prevailing development ideology, a convenient rhetoric was athand. Priority was placed on avoiding political crises in the short run. Policymakers seemedto be little concerned with how policies that were then politically expedient would affectmedium-term or long-term development prospects.

This orientation helps explain the main features of economic policy that I described inChapter 2:

* rapid expansion of government expenditures and proliferation of parastatals.

* strong preference for administrative controls rather than price-based policyinstruments.

* neglect, and even exploitation, of peasant farmers.

* the tendency to ride the terms of trade escalator rather than stabilize themacroeconomic situation.

Many development economists in the 1950s tended to regard the state as an innovator andan entrepreneur. The experience of the Great Depression had reinforced the spirit ofskepticism about markets. In underdeveloped countries particularly, reliance on markets wasconsidered treacherous (Lewis and Kallab 1986, p. 6). These views were adopted by Africanstudents attending Western universities; many of them became senior bureaucrats andministers in postindependence governments. The ideology of state-directed development wasa part of the intellectual environment of the time. This ideology was influential because itfitted so well with the political interests of the rulers.

Government expenditure expanded rapidly in many countries as did government jobs, andmany new projects and programs were initiated. The ostensible justification was to promotesocial and economic development, but these activities also enabled rulers to providepatronage to their political supporters. Attempts by international agencies and somebilateral donors to apply economic criteria in selecting projects had some effect in limitingwasteful expenditures, but many "white elephants" were created nevertheless. Foreign aidfrom bilateral sources financed a substantial part of these public investments. Such aid wasmotivated by diverse objectives, and among them was the desire to promote exports ofcapital goods (Gulhati and Nallari 1988). The discipline of project appraisal fell victim totwo powerful imperatives: for donor governments to help exporters and for recipientgovernments to give patronage to their supporters. The number of parastatals increasedrapidly. Many of them became an instrument of patronage through which jobs and otherperquisites were provided to political supporters of rulers (Hyden 1983).

Apart from the expansion of the public sector (government expenditure and parastatals),Africa has seen a spectacular growth of government regulation of the private sector, largelythrough industrial licenses, import quotas, export licenses, credit controls, and price controls.

6. Fieldhouse (1986) analyzed the extent of continuity or discontinuity in economic policies in Ghana,Nigeria, Kenya, Tanzania, Cote d'lvoire, and Senegal. He concluded that discontinuity between colonialand postindependence policy was most marked in Tanzania and least marked in Kenya, Cote d'lvoire,and Senegal.


To improve efficiency and income distribution through such policies requires a great deal oftimely, reliable information and analytical skill. Most bureaucracies in Africa weredeprived of both, and yet governments chose to adopt these policies rather than to rely onprice-based, market-oriented instruments. This appears to be a paradox when viewed fromthe perspective of economics, but any economic irrationality was overwhelmed by thepolitical attractions of administrative controls. They allow rulers to reward their politicalsupporters and penalize their opponents. Such fine discrimination is not feasible under aregime of market-oriented policies. Furthermore, controls generate rents that can be sharedamong politicians, bureaucrats, and their favored clients.

A prominent feature of economic policy during the 1970s was exploitation of peasantfarmers. This can be explained to some extent by the development ideology popular in the1950s and early 1960s; it emphasized industrialization based on aggressive importsubstitution (almost at any cost). However, these policies were retained in Sub-SaharanAfrica long after the demise of this doctrine in academic circles and in policies of otherLDCs. According to Bates (1981, p. 121), the real architect of the typical African policybundle was a "development coalition" of rulers, bureaucrats, urban workers, localindustrialists (occasionally multinational corporations as well), and commercial farmers.Industrialists and urban workers derived benefits from protection of local industries againstimports and from the supply of low-cost food. Commercial farmers got benefits throughagricultural input subsidies on fertilizers and equipment (which were not used by peasants toany large extent). The bureaucrats and politicians occupied positions of power and shared inthe rents generated by the system of control. The ruler appropriated material benefits fromthis policy framework through favorable treatment of his own agricultural, industrial, andother business assets. Much more fundamentally, however, he was assured of politicalsurvival through the support of the coalition.

The entire policy edifice rested on denying smallholder peasant farmers their rightfulshare in the economic pie. Smallholders found it difficult to organize politically becausethey were numerically large, politically diverse, and geographically scattered. The costs oflobbying were high, and there was the incentive to free-ride. Commercial farmers did notprovide leadership because they were seduced by input subsidies, and many had economicinterests in urban activities. Parties or factions that tried to mobilize peasants into interestgroups were repressed by the government.

Finally, SSA economies had to cope with a substantial measure of instability as a resultof marked fluctuations in international prices of primary export commodities and therefore ofthe terms of trade. Fiscal and monetary policies could have been used to offset thisinstability to some extent. It is possible to accumulate foreign exchange reserves duringperiods in which the terms of trade are favorable and to draw them down when the terms oftrade deteriorate. The pursuit of such a contracyclical policy requires, however, an extendedplanning horizon for economic policy, an ability to learn from past experience, strong coreeconomic agencies (for example, Ministry of Finance and the Central Bank) that can resistpressures to raise outlays during boom periods, and a fair measure of technical sophisticationin macroeconomic management. These conditions were not fulfilled in most SSA countries,which ended up by riding the terms of trade escalator up and down. In fact, several countriesincreased the amplitude of the fluctuations by borrowing abroad during the upswing andpaying higher debt service charges during the downswing. From a political standpoint, itwas much easier to ride up the escalator during boom times than it was to descend during thedownswing. Quite frequently, the latter involved scaling back public expenditures, whichwas difficult to do if expansion during the boom had taken the form of new government jobsor new entitlements. Such inflexibility, the "ratchet effect" of consumption theory, was aprominent feature of public expenditure in Africa, leading to recurring budget deficits andfurther external borrowing during the downswing.

Some of the sample countries had in common certain economic policies as well as politicalparameters. But the relationship between economics and politics is not a straightforwardone. Botswana and Malawi assigned a high weight to economic growth and a low weight to


income redistribution. Both pursued conservative fiscal and monetary policies, although thisstance was adhered to much more consistently in Botswana than in Malawi.7 Both managedto maintain relatively realistic exchange rates. Both avoided a proliferation of governmentinterventions, unrealistic wage hikes, and indiscriminate, high-cost industrialization. Bothavoided premature Africanization of their civil services. These similaritiesnotwithstanding, the two countries were far apart in terms of one key political dimension.Malawi is a strong autocracy, while Botswana is a consultative democracy. The commondenominator of economic policies was that political authorities in both countries had similareconomic interests (livestock in Botswana and tobacco in Malawi) and a nonideologicalapproach to economic issues. Tanzanian and Zambian economic policies offered a sharpcontrast. Government played a very large role in both cases. Nationalization of foreign-owned assets and indigenization of civil service were pursued vigorously. Agriculture wasneglected, and industry was given heavy protection against imports. Nyerere and Kaundashared a certain ideological orientation that affected their economic policies.

Policy Implementation

Smooth implementation of policy and programs requires many preconditions: (1)agreement on objectives; (2) high-quality decisions backed up by valid theory and realisticassessment of resource costs and availability; (3) simplicity of measures to be implemented sothat only a limited number of administrative agencies are involved and so that tasksinvolved can be specified in detail; (4) good communication, coordination, and obedience tohierarchical authority (Cleaves 1980; Jackson 1983). Implementation is rarely smoothbecause few if any countries can meet this demanding list of preconditions. However,implementation failures are much more frequent and much more serious in SSA countries thanin most other parts of the world.

First, in many Sub-Saharan African countries there is much less than full agreement onpolicy or program goals, except at a very general level. The goals of economic policy are notvery different from those found elsewhere. Rapid economic growth, reduction of inequality,poverty alleviation, and financial balance are prominent among the objectives listed indevelopment plans and similar documents. However, the operational significance of suchstatements is extremely limited. Such plans have a tenuous relationship with detailedbudget proposals and even a more remote connection with actual developments in publicfinance. Actual policies and practices relating to, for example, the exchange rate, foreignexchange licensing, import tariffs, producer prices, and parastatals are governed by goalsthat are seldom articulated clearly. Analyses of the impact of such polices frequently revealtheir internal inconsistency and short-term orientation. These findings are scarcely surprisinggiven the nature of the typical polity and the decisionmaking process in Sub-Saharancountries. Little effort is made to build up a genuine consensus among the parties inside andoutside government affected by a policy change. There is little public debate and little

7. Botswana has avoided macroeconomic mismanagement so far. It built up government cashbalances and foreign exchange reserves during the good years. It prevented a real appreciation of itsexchange rate. It restrained wages for macroeconomic reasons. Harvey (1987) advances the followingfactors in explaining the Botswana case:

* Senior expatriate advisers set up a framework for macro-policy. They conducted seminars andbriefings for the cabinet, senior civil servants and members of parliament.... Advisers acquiredpolitical sensitivity, politicians acquired familiarity with some of the technical issues and a lot ofthe statistical work was done before the crisis erupted in the early 1980s.

* Full control over macro-policy was acquired only in 1976-ten years after independence.Meanwhile, experience of other African countries was being distilled by Botswana Ministers andbureaucrats.

* Top leaders were determined to use expert advice to manage their affairs. There was remarkablestability. During the first two decades of independence, there were only two Finance Ministers andonly four permanent secretaries in that ministry.

34 7he Making of Economic Policy in Africa

consultation. Decisions are made on a piecemeal basis, and there is limited awareness of thelogical linkages among them.

While formal policy decisions can be taken by a small group or even by the individualruler, implementation of many policies (such as foreign exchange licensing, agriculturalmarketing, or price controls) requires the collaboration of many individuals and agencieswith diverse personal and organizational interests. It is all too likely that goals driving thepolicy decision will not be congruent with many of these interests and a number of conflictswill ensue. Even when resistance does not take an overt form, opposing forces can adoptobfuscatory and dilatory tactics. Furthermore, since many policy decisions are fairly generaland must be elaborated and detailed by the bureaucracy, there is plenty of room for turningthings around during the implementation phase (Grindle 1980).

Policy and program decisions were much affected by explosive expectations unleashed bypolitical independence. They also reflected the lack of experience of the new leaders.Ambitious targets were adopted that bore little relation to the availability of financial orhuman resources. Leaders made little attempt to educate the public in the realities of theeconomic situation (Ojuka undated, p. 241). Some of the adopted policies were complex (forexample, the rapid expansion of the public sector, nationalization of foreign firms,establishment of new parastatals, integrated rural development schemes, and licensing offoreign exchange), and they aimed at changing the structure of the economy. Rothchild andCurry (1978) classified Sub-Saharan countries into those pursuing accommodation strategies,reorganization strategies, and transformation strategies. Accommodation strategiesemphasized production goals and accepted the existing international and domesticinstitutional framework. Such strategies were the least demanding in political andadministrative terms. Reorganization strategies and transformation strategies visualized arestructuring of existing institutions and a large role for govemment in economic development.Transformation strategies were even more radical than reorganization strategies and imposeda maximum political and administrative burden.

Finally, policy and program implementation in SSA countries was stymied by adversechanges in public administration during the 1970s. Efforts were made to transform thecolonial bureaucratic machine with its focus on law and order into "developmentadministration" through foreign-financed technical assistance, but this was unsuccessful(Hyden 1983). Instead, the pervasive political trend toward personal rule andpatrimonialism undermined bureaucratic norms and principles. Furthermore, the behavior ofpublic officials in Africa was influenced by both their civic duties and their obligations tothe primordial public-groups connected by blood relationships, kinship, and tribalaffiliation. According to Ekeh (1975, p. 108), "the unwritten law of the dialectics is that itis legitimate to rob the civic public in order to strengthen the primordial public." Suchtensions are not peculiar to Sub-Saharan Africa, of course. They exist everywhere. Whatperhaps distinguishes Sub-Saharan Africa is the extent of the tilt away from civic duties.Some observers believe that the state elite has used government machinery to become adominant class. Given their power to allocate scarce resources, bureaucrats not only canfurther their own business or agricultural interests, but they also can grant favors to localindustrialists or commercial farmers.8

The bureaucracy grew very rapidly in the postindependence period. According toAbernethy (1988), public sector employment in Sub-Saharan countries rose from 3.8 million in1960 to about 10 million in 1980, or from 40 to 45 percent of nonagricultural wage employment

8. Adedeji (1986, p. 111) states, "Indeed, the poor economic management of our economies has put aquestion mark on the ability of the public sector to play the kind of dynamic role in the developmentprocess called for by the [Monrovia] Strategy and the ILagos] Plan... Inefficency has become the normin our public service... It seems that within the first quarter of a century after independence Africanpublic services have lost their mission... and have become a self perpetuating, self-centered and self-aggrandizing cabal." Adedeji has been the secretary general of the Economic Commission for Africa(ECA) for many years. He is a distinguished professor of public administration.


to 50 to 55 percent. Corresponding figures in 1980 are 36 percent in Asia, 27 percent in LatinAmerica, and 24 percent in OECD countries.9 Furthermore, the average pay of governmentworkers, relative to GDP per capita, was much higher in Africa (6:1) than in Asia (2.9:1),Latin America (2.9:1), or the OECD countries (1.7:1).

In the classification of Rothchild and Curry (1978), Zambia pursued "transformation"strategies. Its record of policy implementation was flawed (see Chapter 5). Relativelyspeaking, Malawi's implementation failures were less serious than in Sub-Saharan Africagenerally. Mauritius's and Botswana's capacity to implement plans and policies was verygood in most areas. Several factors contributed to this outcome. Decisions in these twocountries were consensual and pragmatic. Government interventions were kept to a feasibleset. For example, the number of parastatals were not allowed to proliferate as happenedelsewhere. Discipline was exercised by the Bostwana government in screening projects, and"white elephants" were avoided. Planners insisted that foreign donors adhere to thepriorities defined by the Botswana government, and they got their way. The quality of theBostwana civil service was maintained at a high level by avoiding precipitate localization,providing high compensation, and maintaining well-defined lines of authority andaccountability (Raphaeli, Roumani, and Mackellar 1985; Picard 1987). Botswana's recordwas far from perfect, however. The government failed to carry out its policy of ruraldevelopment largely because political commitment was weak in this area.

Political trends clearly influenced economic policy decisions and their implementation.The spread of "personal rule" was a major determinant of the strong trend toward statismand the growing importance of administrative controls. It was difficult to comprehend theeconomic rationale of many of these policies, but their political rationale was obvious. Itwas also clear that "personal rule" undermined the professionalism of the civil service.Statism took hold at a time when public administrators were under considerable pressure tobecome the conduit for patron-client relations.

Policy everywhere reflects a combination of technocratic, political, and organizationalconsiderations. In Africa this balance is tilted much more against the technocratic element,and the political aspect is much more influenced by patron-client relations than in manyother regions.

9. These comparisons are influenced by the fact that the denominator (that is, nonagricultural wageemployment) is smaller in the African region than in other country groups.

4POLICY EXPERIENCE IN THE 1980s

Judged in terms of real economic growth or the financial situation, the 1980-86 period wasdisastrous for most sample countries. This was the result of three factors. First, theinternational economic environment was unusually hostile. Second, weaknesses in economicpolicy continued to impede macroeconomic management and the efficient allocation ofresources (see Chapters 2 and 3). Although many countries signed agreements with the IMFand the World Bank to improve these policies, only two (Mauritius and Malawi) sustainedintensive reforms. Several other countries carried out reforms of medium intensity for shortperiods. Third, while reforms had some positive impact, the magnitude of improvements ineconomic performance was circumscribed by the unfavorable initial conditions of most samplecountries and by imperfections in the design of new policies.

The chapter begins with an outline of the economic and financial outcome in samplecountries during the 1980-86 period. Next I review the incidence and size of exogenous shocksaffecting individual countries. This is followed by a brief sketch of policy responses in ninecountries with an emphasis on the relative intensity and continuity of reform efforts. Thesepolicy profiles merely whet the appetite, and many readers will be left with unansweredquestions about significant details. The main purpose served by these short sketches is toreinforce the classification of sample countries presented in Table 4.5. Readers who arealready familiar with some of these countries may find this brief treatment useful. Othersmay turn to the references at the end of the study for a full analysis of policy developmentsin individual countries. Three countries-Zambia, Malawi, and Mauritius-are given a fulleranalysis in Chapter 5. It should be recognized, however, that this classification of countriesin Table 4.5 remains subjective, and it is not even possible to define intensity and continuityvery precisely. Nevertheless, these terms are suggestive, and I hope that readers will acceptthat my classification of countries is at least plausible.

Macroeconomic Changes

Most countries in the sample recorded a slower increase in output from 1979 to 1986 thanfrom 1973 to 1979 (see Tables 2.1 and 4.1). Remember that during the latter period there wasconsiderable deceleration. Since population was rising at an annual rate of approximately 3percent, GDP per capita declined during the 1980s for most countries. The only exception wasMauritius. The fall in per capita consumption was very sharp in Madagascar, the Sudan,Zambia, Zaire, and Kenya. Furthermore, the volume of investment declined sharply in fourcases and stagnated in two others. The volume of imports shrunk in nine countries.Marked contraction in the real economy during the 1980s was accompanied by severefinancial disequilibrium in many countries. The average external deficit on current accountduring the 1980s was much higher than in the 1973-79 period in the Sudan, Zambia,Madagascar, and Ethiopia (see Tables 2.4 and 4.2). It remained at 5 percent of GDP orhigher throughout the sample, except for Tanzania and Uganda. Balance of paymentpressures in most sample countries were reflected in rising debt service and the drawing downof external reserves. The burden of external debt had become much heavier in the 1980s thanin the preceding decade. It was reflected in high debt service ratios and in repeatedrescheduling by the Paris Club and other creditor groups. Debt service ratios were low in

37


Table 4.1 Indices of Economic Change, 1979-86(average annual percent growth rate, 1980 prices)

Total GovernmentCountry GDP Consumption investment GDEa expenditureb Exportsb Importsb

Zimbabwe 2.9 1.4 -0.4 2.3 4.9 1.0 -2.5Mauritius 3.5 1.9 3.6 1.5 1.0 4.3 -0.3Kenya 2.9 0.1 -3.1 0.5 3.6 -1.4 4.4Somalia 0.8 0.3 2.6 0.8 5.4 -16.3 -3.6Uganda 1.5C -2.4 d 6 .5 d 41d 2. 0.3 3.5Malawi 2.0 15 -7.7 -0.8 0.8 2.2 -5.4Tanzania 1.2 3.1 4.6 1.7 -10.6 -8.4f -1.4fZaire 1.1 -0.8 1.3 -0.8 3.8 -1.6 2.6Ethiopia 0.9 1.9 4.1 2.4 53 2.8 6.3Zambia 0.4 0.1 4.7 -1.0 0.2e -6.6 -3.7Sudan 0.1 0.1 0.9 0.1 2.6 -1.1 -4.4Madagascar -0.8 -1.7 -7.9 -2.1 -7.6d 5.5 -5.5

Note: Least square growth rates were calculated.

a. The data are not consistent with that for consumption plus total investment owing tosizable statistical discrepancies.

b. Compound growth rates are based on three-year averages for both the initial and terminalperiod.

c. 1980-86.

d. 1979-83.

e. 1979-85.

f. Goods and nonfactor services.

Source: The data presented in the first four columns were derived from national accounts innational currencies. Data on exports and imports are from UNCTAD. Government expenditurefigures are from IMF, Government Finance Statistics.

Sudan and Zambia in 1986 only because of the large accumulation of arrears.1 Mauritius wasthe only country in the sample with a manageable debt burden throughout the 1980s.Average proportions of external reserves to imports remained lower than 20 percent in tencases. Table 4.3 shows available data on budget deficits (government net borrowing fromdomestic banks) and inflation. It is clear that double-digit inflation persisted in all samplecountries except Ethiopia. The average annual rate of inflation exceeded 20 percent in sixcases.

It ought to be possible to analyze these economic and financial outcome indicators in theframework of a general equilibrium model that takes account of (a) initial conditions at theend of the 1970s, (b) exogenous shocks during the 1980-86 period, and (c) key policy variables.

1. Sudan's scheduled debt service in 1986 was $899 million, but it paid only $55 million while arrearsincreased correspondingly. Similarly, Madagascar paid only $113 million in 1986 in comparison with the$241 million that was falling due before rescheduling action by creditors. Debt service ratios convey amisleading impression of the debt burden because repeated rescheduling by the Paris Club an othercreditor groups postpones the date when debt service falls due.

Policy Experience in the 1980s 39

Table 4.2 Indices of External Financial Disequilibrium, 1980-86

External deficita Debt service ratiob External reservesC

Country Average Peak 1986 Average Peak 1986 Average Trough 1986

Zaire 6 7(1986) 7 14 19(1986) 19R 24 11(1982) 33

Sudan 11 18(1982) 10 12 18(1982) 7R 2 1(1981) 6

Zambia 13 21(1986) 21 12 24(1981) 2R 10 5(1981) 10

Madagascar 11 17(1980) 9 19 31(1985) 30R 8 2(1980) 17

Uganda 4 7(1980) 1 15 21(1981) 7R 16 1(1980) 8

Tanzania 4 8(1980) 4 13 16(1986) 16R 3 0(1982) 7

Malawi 9 21(1980) 7 27 41(1986) 41R 13 5(1983) 10

Zimbabwe 5 10(1982) S 17 33(1983) 23 11 6(1984) 11

Ethiopia 7 9(1985) 6 14 26(1986) 26 19 6(1984) 24

Somalia 5d 8(1982) 14e 21 44(1986) 44R 3 2(1984) 3

Mauritius 6 13(1981) 7 11 16(1983) 7 9 4(1983) 20

Kenya 5 12(1980) 1 20 26(1985) 23 21 11(1981) 25

a. Deficit on current account of the balance of payments excluding official transfers aspercentage of GDP. S = current account surplus.

b. Actual debt service payments as percentage of exports of goods and services. R =

rescheduling of debt one or more times during 1980 to 1986.

c. Total reserves (gold valued at SDR 35 per ounce) as percentage of imports of goods andservices.

d. Average 1980-85.

e. 1985.


On the basis of robust econometric relationships between relevant outcome indicators and theset of the above three independent variables, one could not only explain what actuallyhappened, but also carry out simulations of the counterfactual. Unfortunately, this analysisis beyond my reach in this study owing to the unavailability of reliable data and theabsence of well-tested models. Some authors have carried out crosscountry analysis comparingeconomic outcomes in countries undertaking reforms with a control group consisting of countriesnot undertaking reforms. One such study based on a sample of forty-eight countries concluded"that the average performance of AL countries [those carrying out reforms] was better thanthat of NAL countries [those not carrying out reforms] (World Bank 1988, pp. 2.3).Alternatively, outcome indicators before reform could be compared with similar indicatorsafter reforms. Such studies are suggestive, but they are scarcely conclusive. In view of theselimitations, I prefer to supplement the above analysis of what happened to the samplecountries in the 1980s with a separate assessment of exogenous shocks and policydevelopments.


Table 4.3 Indices of Internal Financial Disequilibrium, 1980-86

Budget deficita Inflationb

Country Average Peak 1985 Average Peak 1986

Uganda 2 5(1981) 3 74 n.a. 169Zaire 3c 6(1982) od 45 77(1983) 47Sudan 3 5(1985) 5 30 45(1985) 26Ethiopia 5 9(1982) 4 5 19(1985) -10Mauritius 4 7(1983) S 13 42(1980) 2Somalia 2 4(1984) 1 47 59(1980) 36Madagascar 5 11(1980) 1 19 31(1981) 15Zambia 5 16(1982) 6 24 52(1986) 52Kenya 2e 3(1982) n.a. 12 20(1982) 4Zimbabwe 3 3(1980) 3 14 23(1983) 14Malawi 4 9(1981) 2 14 20(1984) 14Tanzania 6f 7(1982) n.a. 31 35(1984) 32

n.a. not available.

a. Government net borrowing from domestic banks as percentage of GDP. S = budget surplusdefined in the same way.b. Average annual increase in cost of living index.c. Average 1980-84.d. 1984.

e. On a gross basis, average 1980-83.f. Average 1980-83.


Hostile International Environment

An updating of the analysis of exogenous shocks by Gulhati and Yalamanchili (1988)leads to four conclusions (see Table 4.4). First, nine sample countries lost more than 1 percentof gross domestic income (GDY) owing to a deterioration in their terms of trade in 1979 to1981, compared with the preceding three years. The loss was more than 3 percent in threecases. Further worsening occurred in eight countries from 1981 to 1983 compared with 1979 to1981. The following three years, 1983 to 1985, saw some improvement in the terms of trade ofa large proportion of sample countries, but the average index remained substantially belowthat of 1980 in Kenya, Mauritius, Zaire, and Zambia.

Second, foreign concessional aid rose as a share of GDY in the entire sample during the1979-81 period, compared with the preceding three years. The 1981-83 period saw furtherimprovements in this ratio, except in three cases. However, there was a widespread setbackduring the 1983-85 period; the ratio of aid to GDY fell in seven cases.Third, loans on conventional terms from commercial banks, suppliers' credit agencies, and theInternational Bank for Reconstruction and Development (IBRD) yielded an increasing netresource transfer to five sample countries from 1979 to 1981, to three countries from 1981 to


Table 4.4 Impact of Exogenous Shocks During the 1980s

Terms of trade ODAa Loansb Crop outputc

Country 1979- 1981- 1983- 1979- 1981- 1983- 1979- 1981- 1983- 1979 1981- 1983-81 83 85 81 83 85 81 83 85 81 83 85

Mauritius -5.9 -1.5 1.0 0.9 0.8 -1.4 -0.6 -4.5 0.6 -1.4 1.0 0.3

Somalia -2.5 -0.7 -1.7 5.3 6.0 -0.8 -0.1 0.9 -1.7 3.3 6.4 13.5Malawi -4.9 1.0 -0.5 1.3 -1.1 -0.2 -2.1 -2.5 -0.9 0.7 3.9 1.0Ethiopia -2.6 -1.2 0.1 0.7 0.4 5.5 0.3 0.2 -0.1 4.6 2.5 -4.6Kenya -2.5 -0.7 0.6 1.2 0.5 -0.6 -1.0 -2.0 -0.6 -0.2 2.3 1.6Tanzania -2.1 -0.5 0.7 2.2 0.4 -2.5 0.1 -0.3 -0.8 4.3 -1.2 0.3

Madagascar -1.8 -1.0 1.0 2.3 2.4 -1.6 1.6 -3.5 -1.1 -0.1 1.4 2.6Zaire -1.6 2.7 -1.0 0.6 -1.1 -0.4 -4.0 -1.0 -0.4 1.4 2.6 1.7Sudan -0.4 -0.4 1.0 2.0 1.9 1.9 0.8 -0.4 -1.1 -1.2 -1.5 -0.4Zimbabwe 1.1 2.5 0.5 1.8 1.3 0.6 -1.4 1.8 4.0 1.9 -02 0.8Zambia -0.2 -2.2 6.6 0.7 -0.5 0.1 1.1 -0.2 -0.1 -1.4 0.1 1.6Uganda -3.4 1.5 n.a. n.a. n.a. n.a. n.a. n.a. n.a. -2.9 7.2 12.3

n.a. not available

Note: Comparison is made with the average of the preceding three years. The change in each variable is expressedas a proportion of GDY during the current three-year period. The change in offical development assistance (ODA)is expressed in 1980 prices using a special deflator for this variable developed by the Organization for EconomicCooperation and Development (OECD). A United Nations index for unit value of manufactured goods is used todeflate loans. An index by the U.N. Food and Agriculture Organization (FAO) of physical output is used for cropoutput. For details of the approach underlying the calculations in the table, see Gulhati and Yalamanchlii (1988).

a. Net disbursements.

b. Net resource transfer on account of loans from IBRD, suppliers' credit agencies, and conmnercial banks.

c. Food and nonfood crops.


1983, and to only one country from 1983 to 1985. In fact, this transfer had become negative forseven borrowers in the 1983-85 period. There was a sharp decline in the flow of newcommitments and gross disbursement of these loans. Fourth, a number of countries sufferedadverse shocks on account of changing weather conditions. Fall in crop output as a proportionof GDY was almost 3 percent in Uganda from 1979 to 1981 compared with the preceding thereyears. The corresponding loss in Ethiopia from 1983 to 1985 was even larger. My method ofcalculation tends to understate these shocks since I use three-year averages, while manyweather-induced disturbances (droughts, floods, cyclones) reverse themselves after one or twocrop seasons.

Different Policy Responses

Enlightened policymakers can be expected to adopt a proactive posture in tacklingproblems. They can exercise initiative in "choosing" a policy agenda, preparing a policyprogram, and implementing new policies in a relatively favorable environment (Hirschman1981, p. 146). There are clear advantages in undertaking reforms when the economy is strongand allows considerable room for maneuver in the event of setbacks. Furthermore, neweconomic policies have a better chance of succeeding if bureaucratic and political factors donot necessitate departures from the "rational-actor model" (Allison 1971). The economic


policy process in Sub-Saharan Africa did not fit this model during the late 1960s and 1970s(see Chapter 3). Not surprisingly, policy developments during the 1980s also did not fit thismodel. Reforms were "reactive" rather than "chosen." They were undertaken in the contextof deteriorating economic and financial conditions, thereby reducing the room for maneuverand making policy change more risky. The extent of the departure from the rational-actormodel, necessitated by political and bureaucratic factors, varied a great deal from country tocountry.

Most Severely Affected Countries

A plausible hypothesis drawn from the rational-actor model would be that very strongreform efforts would be made by countries most severely affected by the economic andfinancial crisis. A look at Table 4.5 does not confirm this hypothesis. A number of thesecountries did not develop a meaningful reform program until 1983 or 1984. Their first reactionwas either to deny that there was a serious problem or to externalize it by attributingundeniable difficulties to unfavorable changes in their terms of trade or, more broadly, tothe existing international econornic order. There was considerable reluctance to accept thenotion that domestic policies might be responsible for some of the problems. Time was losthoping that the terms of trade would improve or that donors would organize a rescue. Suchreactions to adversity were fairly common worldwide (Dror 1986). Uganda and Zambiaundertook intensive reforms for three years, but these programs could not be sustained. Zaireand Madagascar signed many policy agreements, but the record of implementation was acheckered one. In particular, a number of agreements with the IMF could not be completed.The following paragraphs sketch out these events very briefly in each country.

Table 4.5 Intensity of Policy Responses, 1980-86

Country 1980 1981 1982 1983 1984 1985 1986

Most severely affected by crisis

Uganda L H H H L L LZambia L L L M H H HZaire L L L M H H MMadagascar L L L M M M M

Substantially affected by crisis

Sudan M M M M L L LSomalia L M M M L M MTanzania L L L L L L MZimbabwe L L L L L L LEthiopia L L L L L L L

Moderately affected by crisis

Malawi M H H H H H HMauritius M H H H H H HKenya M M M M M M M

Note: H = high policy reform intensity; M = medium policy reform intensity; L = lowpolicy reform intensity.


Uganda adopted intensive reforms in the early 1980s (see Table 4.5). This major shiftoccurred after almost a decade of protracted economic decline, severe loss of human capital,and chronic disintegration of institutions under Idi Amin, who was classified as a "tyrant"by Jackson and Rosberg (1982), as noted in Chapter 3. His ouster was followed by more than ayear of political instability and then the general election in December 1980 of PresidentMilton Obote as the head of the Uganda People's Congress. Although the new governmenthad an electoral mandate (despite accusations by rival parties of serious rigging of votes), itfaced a very difficult political situation (Kamuntu 1987). The historical schism between theBuganda (a major tribe) and the rest of the country persisted. There was a particulargrievance against Obote who had established a republican constitution in the late 1960safter sending the Buganda king into exile. Furthermore, the new government in 1980 did nothave decisive civilian control over the army, and there was considerable insecurity. Obotewould probably fall under the category of a "prince" in the Jackson and Rosberg typology.

In this extremely fragile situation, economic reforms were launched in Uganda on thepremise that economic recovery would lead to political improvements and a reduction ofinsecurity (see Table 4.6). The government was prepared to take the big risks implicit in the"shock treatment approach" because it aimed at an economic turnaround in not more thanthree years (Kamuntu 1987). Stiff stabilization and structural adjustment measures wereadopted, but the policy gamble did not pay off. The reform effort was derailed toward theend of 1984. In tracing the reasons for this failure, Kamuntu draws attention to a number ofimperfections in the technical design of the reforms, to serious administrative weaknesses inimplementing policy decisions, to the inability of the IMF and the World Bank to mobilizesufficient external resources, and to the inflexibility of the International Coffee Agreementwith respect to Uganda's export quota. Finally, Kamuntu points out that although thereforms contributed to a number of positive outcomes in output, exports, financial balances,and inflation, they did not trigger a decisive economic revival. Meanwhile, the securitysituation worsened, and politicians became infected with election fever. A group oftechnocrats (headed by Ephraim Kamuntu, who had exercised very considerable influence onpolicy in the early years of the reform) was increasingly sidelined as elections approached.There was a military coup in July 1985 followed by a civil war. The new Musevenigovernment took time deciding what approach to follow. Initially, there was considerableskepticism regarding the prescriptions of the IMF and the World Bank. An attempt wasmade to design an alternative solution, but finally in 1987 relations with the IMF wereresumed.

Table 4.6 Uganda: Chronology of Selected Policy Events, 1980-86

Month/Year Event

January 1980 Macro policy agreement with IMF; completed December 1980.June 1981 Macro policy agreement with IMF; completed June 1982.November 1981 Debt rescheduled by Paris Club.May 1982 Consultative Group of Aid Donors reviews economic policies.August 1982 Macro policy agreement with IMF; completed October 1983.December 1982 Debt rescheduled by Paris Club.February 1983 Agricultural policy agreement with IDA.September 1983 Macro policy agreement with IMF; implementation was incomplete.January 1984 Consultative Group of Aid Donors reviews economic policies.July 1985 Military coup against OboteJanuary 1986 End of civil war; Museveni government assumes power.


GDP in real terms rose by 19 percent during 1980 to 1983 from a very low level but thensuffered a serious setback. Government borrowing from local banks was reduced from 4.6percent of GDP in 1981 to zero in 1984, but there was a relapse subsequently. The rate ofinflation too was brought under control for a brief spell before accelerating once again.

Zaire's economic difficulties were the combined result of severe exogenous shocks andacute distortions in policies (see Table 2.11). Its story of attempted reforms has received agreat deal of attention by political scientists (Callaghy 1986a; Young 1985; and Nzongola-Ntalaja 1986). It is a complicated story with many nuances, and this brief summary cannotdo justice to the rich detail presented by various authors. The main point, however, is thelong history during the 1970s and early 1980s of short-lived stabilization programs (seeTable 4.7). In each case the Mobutu regime managed to sign agreements with the IMF (andwith its external creditors) based on promises to reform economic management. Theseagreements secured external resources, but pledges to improve management were repeatedlybreached by the Zaire authorities.

Table 4.7 Zaire: Chronology of Selected Policy Events, 1979-86

Month/Year Event

August 1979 Macro policy agreement with IMF; implementation was incomplete.April 1980 Debt rescheduled by commercial banks.June 1981 Macro policy agreement with IMF; implementation was incomplete.July 1981 Debt service rescheduled by Paris Club.January 1983 Debt service rescheduled by commercial banks.December 1983 Debt service rescheduled by Paris Club.December 1983 Macro policy agreement with IMF; implementation was incomplete.June 1984 Debt service rescheduled by commercial banks.April 1985 Macro policy agreement with IMF; completed April 1986.May 1985 Debt service rescheduled by commercial banks.September 1985 Debt service rescheduled by Paris Club.May 1986 Debt service rescheduled by commercial banks.May 1986 Macro policy agreement with IMF; implementation was incomplete.June 1986 Industrial policy agreement with IDA.June 1986 Debt service rescheduled by commercial banks.

Callaghy (1986a, p. 312) argues that the implementation of IMF conditionality wouldhave undermined the "patrimonial regime." (Callaghy describes Mobutu as a "presidentialmonarch," which is roughly equivalent to an "autocratic ruler" in the framework of Jacksonand Rosberg). Mobutu could avoid genuine reforms without losing external support becausesome Western powers were afraid of what might happen in political-strategic terms if therewas a breakdown of the current regime in Zaire. "The desire for and commitment to changeand reform varies from [Westemr government to government, and ebbs and flows within eachWestern government as administrators and policy-makers change. Mobutu has taken goodadvantage of these changes and lapses of attention" (Callaghy 1986a, p. 324). In 1978, theZaire government agreed to install expatriate expert teams in key economic agencies tofacilitate the implementation of policy agreements with the IMF.7 Many observers weretroubled by this experiment, which seemed extremely intrusive, but it failed to deliver thegoods, in any event. Callaghy (1986a, p. 322) concluded, "The valiant efforts of the various

Z Ghosh (1986, p. 464) states, "In early 1979, there was not a single major area of the economy whichwas not in the hand of foreigners-a feature which caused some commentators to characterize Zaire 'asthe model neocolony.' "


internationally-sponsored teams at the Bank of Zaire, the office of Debt Management,Customs, Finance, and Planning have been distinctly limited in their impact. The maneuversof Mobutu and his political aristocracy to detour the controls have been creative, persistentand, to a substantial degree, successful."

When Zaire mounted a macro policy program once again in 1983, there were good groundsfor healthy skepticism. The initial results, however, were fairly impressive. Manyagricultural marketing parastatals were dismantled. Most prices, including agriculturalprices and interest rates, were decontrolled. Government-regulated prices of petroleumproducts were raised sharply. An interbank auction for foreign exchange was established inSeptember 1983, and some measures aimed at simplification of the trade regime were taken.The state mineral marketing monopoly, whose operations had raised many eyebrows, wasabolished. There was evidence of improved budgetary control and some reduction in thenumber of redundant government workers. These developments built up an optimistic mood insome quarters. According to Young (1985, p. 5), "The regime seemed to feel more politicallysecure than in the past, thus altering its own calculus of risk and benefit in appraisinginternationally prescribed austerity and structural adjustment programs." Other favorablefactors mentioned by Young were a relative lack of ideological constraints, an improvedmilitary situation (following a lowering of the Soviet posture vis-a-vis Angola and arapprochement with Israel), and an increased interest on Mobutu's part to secure economicrecovery.

The second half of 1986 witnessed a setback in Zaire, and alarm signals rang once again.Export prices declined as did public revenues, and government borrowing crowded out theprivate sector. IMF credit ceilings were violated. At an October meeting of the Party'sCentral Committee, decisions were taken to reverse major reforms. These reversals weredenied in January 1987, and government negotiated two policy loans from the World Bank in1987. The zig-zag process continues, and there is little stability in the policy climate.

The sharp economic decline that started in the 1970s in Zaire had not been reversed bythe middle of the 1980s. Although total GDP in real terms increased slowly in the 1980scompared with a rapid decline in much of the previous decade, total consumption continuedto diminish. Even in terms of financial stabilization, the record had not improved much. Theaverage rate of inflation was 54 percent between 1973 and 1979 and 45 percent between 1980and 1986. The pressure of debt service was relieved temporarily by four reschedulings in the1970s and another nine in the 1980s.

Next a brief sketch of Madagascar is presented. Its problems were caused by serious policydistortions and the impact of large shocks (see Table 2.11). In the early years afterindependence, the country experimented first with capitalism, then state capitalism, andfinally radical populism. In 1975, Didier Ratsiraka established a regime based on principlesof humanist Marxism, but one which included diverse ideological elements.3 The governmentwas a functioning multiparty system and a hybrid civil-military regime. According to Covell(1987, p. 162), "The coalition was assembled in part in spite of the adoption of Marxism,through modification in the ideology, the offering of material inducement and the threat offorce." This ambiguity in the revolution makes it difficult to classify Ratsiraka in terms ofthe Jackson and Rosberg typology. This very same ambiguity provides an insight into policydevelopments that occurred in the 1980s.

The Malagasy government started out by undertaking three stabilization programs withthe IMF that had to be aborted (see Table 4.8). Starting in 1982, the president changed theleadership structure progressively and brought into the policy circle young technocraticministers and advisors who believed in carrying out a serious review of economic policy. Thefourth stabilization program made some headway, and this was followed by policyagreements with the World Bank on some industrial and agricultural issues. The government

3. According to Young (1982, p. 50), "The sincerity and depth of the ideological engagement of theleadership is doubted both by non-socialist foreign observers and the doctrinally pure scientific socialistintellectuals at home."


succeeded in reducing the budget and external deficits, restraining inflation, liberalizinginterest rates, depreciating the real exchange rate, phasing out some import restrictions, andrationalizing the export regime. The marketing of rice was liberalized. These reforms wereof medium intensity (see Table 4.5). The political setting did not permit a sharp and rapidpolicy turnaround. At the end of 1986, a great deal remained to be done. It is not easy tointerpret recent policy changes in terms of Malgash humanist Marxism. Covell suggests that"Marxism is, therefore, seen largely as a strategy for conducting a strong nation with a strongstate. It is the instrumental view of the ideology, as well as brute necessity, that explainsthe modifications in policy undertaken since 1981" (1987, p. 158). It is difficult to say towhat extent Ratsiraka and his cohorts are genuinely convinced that new policy directionsare right. A big gulf separates policy decisions and policy implementation.

Table 4.8 Madagascar: Chronology of Selected Policy Events, 1980-86

Month/Year Event

June 1980 Macro policy agreement with IMF; implementation was incomplete.April 1981 Macro policy agreement with IMF; implementation was incomplete.April 1981 Debt service rescheduled by Paris Club.November 1981 Debt service rescheduled by commercial banks.July 1982 Debt service rescheduled by Paris Club.July 1982 Macro policy agreement with IMF; implementation was incomplete.April 1983 Consultative Group of Aid Donors reviews econornic policies.March 1984 Debt service rescheduled by Paris Club.April 1984 Macro policy agreement with IMF; completed March 1985.October 1984 Debt service rescheduled by commercial banks.November 1984 Consultative Group of Aid Donors reviews economic policies.January 1985 Industrial policy agreement with IDA.April 1985 Macro policy agreement with IMF; completed April 1986.May 1985 Debt service rescheduled by Paris Club.May 1985 Debt service rescheduled by commercial banks.September 1986 Macro policy agreement with IMF; completed February 1988.April 1986 Consultative Group of Aid Donors reviews economic policies.May 1986 Agricultural policy agreement with IDA.October 1986 Debt service rescheduled by Paris Club

Government ambivalence regarding some reforms can be illustrated by examining theexperience gained in the course of decontrolling the rice market. The main rice tradingparastatal, Societe Interet Nationale Pour l'Agriculture (SINPA), was restructured in 1983,and the government granted private traders a more prominent role than they had before.The right of government to control these traders was retained, however. After 1983, thegovernment reduced progressively the imports of rice (in agreement with the IMF) toeconomize foreign exchange. Nevertheless, the government continued to try and keepconsumer prices at low levels, to maintain consumption at 400 grams per persons per day, andto adhere to a time phasing of rice imports that ignored the seasonality of local output(Shuttleworth 1989). These policies were inconsistent with the movement toward decontrol.Rice availability dropped sharply, and prices in parallel markets rose markedly during1985-a year of considerable strain.

Real GDP in Madagascar continued to decline in 1983 and 1984 after which there was areversal. Nevertheless, it remained in 1986 far below the peak reached at the outset of thedecade. A measure of stabilization was achieved, although the basic situation in 1986


remained rather fragile. The ratio of actually paid debt service to exports remained at 30percent in 1986 despite eight rescheduling operations during that decade.

Substantially Affected Countries

The experience of "substantially affected countries" was not very different from that ofthe first group, "most severely affected countries." The IMF and the World Bank were fairlyactive in a policy sense in Sudan and Somalia but not on a continuous basis. Programs tendedto get derailed along the way. Governments seemed to follow a "stop-go routine" instead ofmaking a sustained policy effort. In Tanzania and Zimbabwe protracted controversiesbetween governrnents and international agencies tended to hold up progress for long periods.And Ethiopia persisted in its search for a route to adjustment different from thatrecommended by the IMF and the World Bank throughout the 1980-86 period.

Sudan's economic and financial difficulties at the end of the 1970s were largely the resultof acute policy weaknesses and also of some shocks of medium intensity (see Table 2.11). The1973-79 period had witnessed a massive investment and consumption boom, financed at theoutset by oil-exporting Arab governments and later by foreign suppliers' credit agencies andcommercial banks. By 1978 there was a balance of payment crisis, but it took considerabletime for the authorities to come to grips with the facts and their implications. The PlanningMinistry and the Finance Ministry did not see eye to eye, and it was only after much delaythat the two agencies of government finally resolved their differences.

The government was not able to sustain the 1979 macro policy agreement with the IMF forits full duration (see Table 4.9). The follow-up agreement met a similar fate, but the nextagreement was carried out fully. Implementation of the 1984 macro policy agreement wasdisrupted, however, and no further agreements could be negotiated. 4 In February 1986, theSudan was declared ineligible for IMF resources since it was unable to avoid protractedarrears with respect to payment obligations. External debt was rescheduled on eightoccasions. The balance of payment continued to show a large deficit and external reservesremained far too low throughout the 1980s. Little advance was made in controlling inflation.

The International Development Association (IDA) helped the Sudanese governmentdesign an Export Action Program. The Three-Year Public Investment Program was producedfive months after the agreed schedule, and even then it contained projects that IDAregarded as inconsistent with the economic recovery strategy. After some further delay theplan was revised, and controversial projects were excluded. The government could not deliveron its promise to raise revenues from 14 percent to 18 percent of GDP. It established a DebtManagement Unit, but inadequate staffing prevented the unit from functioning effectively.There was considerable success in raising cotton exports (to about 25 percent of total exports)from very low levels reached in 1980-81. Irrigated cotton output increased by 7 percent in1981-82, a further 42 percent in 1982-83, and an additional 3 percent in 1983-84.To a substantial extent, these increases can be attributed to government policy measures suchas abolition of export taxes, replacement of the "joint account" by an "individual account"system on irrigated schemes (which removed the bias against cotton vis-a-vis other crops),and prompt payments to farmers at the end of the season (World Bank 1985, p. 25).

46 According to Brown (1988, p. 76), the negotiation of the 1984 standby was extremely difficult owingto the fact that Sudan could not service payments falling due to the IMF. The U.S. government madeextraordinary efforts to resolve this issue on behalf of the Sudan, and the standby was finally signedafter a delay of several months. Before the Paris Club or the Consultative Group could meet again tofind more money for the Sudan, the government once again fell behind in its payments to the IMF andthe standby was frozen. The U.S. AID then sent a team of Princeton academics to develop an"alternative economic strategy." Brown says that the "U.S. AID was simply seeking an immediate andpolitically expedient solution to the problem of persuading the U.S. Congress and the rest of the donorcommunity to release the aid monies frozen in late 1984, with a view to paving the way for a 1985 IMF-approved salvage operation" (p. 80).


Table 4.9 Sudan: Chronology of Selected Policy Events, 1979-86

Month/Year Event

May 1979 Macro policy agreement with IMF for three years; implementationwas incomplete.

November 1979 Debt service rescheduled by Paris Club.March 1980 Cotton export action program agreed with IDA.November 1981 Debt service rescheduled by commercial banks.February 1982 Macro policy agreement with IMF; implementation was incomplete.March 1982 Debt service rescheduled by commercial banks.March 1982 Debt service rescheduled by Paris Club.February 1983 Debt service rescheduled by Paris ClubFebruary 1983 Macro policy agreement with IMF; completed February 1984.April 1983 Debt service rescheduled by commercial banks.June 1983 Agricultural policy agreement with IDA.May 1984 Debt service rescheduled by Paris Club.June 1984 Macro policy agreement with IMF; implementation was incomplete.April 1985 President Numeiri was deposed.February 1986 Sudan declared ineligible for further loans by IMF.

Unfortunately, a decline in world cotton prices during the 1980s meant that the Sudan couldnot reap the full advantage of its recovery in production. Subsequently, the countryexperienced marketing difficulties: the white fly infestation had made Sudanese cottonsticky.

There was hardly any increase in Sudanese GDP during the 1980s, despite the above-mentioned brief period in which there was a recovery in cotton production. This protractedstagnation should be attributed to the chronic shortage of foreign exchange, to thewidespread impact of the drought (which reduced crop output by more than 10 percent duringthe 1982-84 period compared with the previous three years), and to the inability of thegovernment to pursue forceful and sustained economic policy reforms. I have classified theSudan in the category of countries that made a medium-level policy effort during the 1980-83period, followed by a much weaker program in later years (see Table 4.5).

To understand the causes of the aborted reforms in the Sudan, one must consider, amongother factors, the evolving political scene. President Jaafar Nimeiry came to power in May1969 as a result of a coup d'etat. According to O'Neill (1988, p. 48), the president soonalienated the religious parties of the right and the leftist Sudanese Communist Party (SCP).His most important achievement was the rapprochement with Southern Sudan, which endeda protracted civil war in 1972. This support from the South helped sustain the president inpower. Nimeiry was a "princely ruler" in the Jackson and Rosberg scheme. His room formaneuver in undertaking economic reforms was limited, however. Government decisions towithdraw the wheat subsidy, to raise the petrol price, and to double the sugar price inAugust 1979 precipitated riots and a nationwide strike by railway workers under theinfluence of the SCP. The tenants in the Gezira irrigation scheme protested the higher waterrates imposed by the government.

Meanwhile, disaffection with the Nimeiry regime was growing in the Southern Sudan.This was exacerbated by the decision to pipe crude oil (from a newly discovered deposit inthe South being developed by Chevron) to Port Sudan instead of building a refinery on thesite. Furthermore, the Jonglei Canal (an expensive project in the South to expand theavailability of Nile waters in Egypt and Northern Sudan) appeared to confirm the view ofsouthern observers that the South was being neglected (Ottaway 1987, p. 895). The signing of


a Charter of Integration in 1982 between Egypt and the Sudan led people in the SouthernSudan to fear occupation by northerners (O'Neill 1988, p. 51). The decision to split theSouthern Region into three units in May 1983 proved to be a major debacle. In September1983, the Sharia Law was imposed on the whole country, which caused acute resentment inthe non-Moslem Southern Sudan. Armed opposition broke out, and John Garang establishedthe Sudanese People's Liberation Movement (SPLM). The fighting spread over time, and in1985 work stopped on the Jonglei Canal and the Chevron oil pipeline. Many observersregarded oil exports as a possible way to lessen the balance of payment deficit. Disruption ofwork on the pipeline was a major setback.

President Nimeiry was deposed in April 1985. The precipitating factor seemed to be theprice increases of petrol and bread the previous month, which led to mass demonstrationsand widespread riots. However, a full assessment of the causes of this coup would have toinclude many other factors that had been undermining respect for Nimeiry during the sixteenyears he had presided over state activities. Furthermore, events after Nimeiry's exit do notseem to have provided much relief from the impasse that existed in the Sudan. Electionswere held in April 1986, although the war prevented polling in many southernconstituencies. About 80 percent of the seats in the new assembly were won by parties with adistinctly Islamic character (Ottaway 1987, p. 892). The new government, headed by Sadiqal-Mahdi, was not able to adopt a suitable economic program or to find a solution for theinsurrection in Southern Sudan.T

Somalia is unusual in that most of its citizens belong to one ethnic group. There are manyclans, however, and their conflicts affect political behavior. A coalition government ruledfrom 1960 to 1969, when the army seized power and proclaimed a socialist revolution. Thenew leader, Siad Barre, is an autocratic ruler in terms of the Jackson and Rosberg typology.

Fueled by very large remittances from Somali workers in the Gulf, consumption boomed inthe late 1970s. The financial situation had deteriorated greatly, partly due to a seriousdeterioration in the terms of trade. It was only in 1980, after a substantial delay, that thegovernment decided to stabilize the economy. The IMF standby signed by the government inFebruary 1980 could not be completed, but the following two standbys were fullyimplemented (see Table 4.10). There was substantial improvement in restoring financialbalances. 6 The government terminated its policy of acting as a residual employer of allsecondary school graduates. Government monopoly on grain marketing was abolished.

The expectation was that two successful stabilization programs in Somalia would befollowed by a medium-term structural adjustment program. In this context the governmentprepared a medium-term recovery program consisting of a Public Investment Program, 1984-86,and a phased program of policy reforms in a number of different fields. IDA staff members inthe October 1983 Consultative Group Meeting criticized the public investment program forbeing overly ambitious and for including several doubtful projects. The government respondedby scaling down the size of the investment program and by deleting several projects. The

5. According to Ottaway (1987, p. 891), the government of Sadiq al-Mahdi was "indecisive to the pointof paralysis." Ottaway argues that "the paralysis cannot be attributed primarily to Sadiq's personalshortcomings, but to the nature of the government coalition and the fact that the solutions to thevarious problems may well be inherently contradictory and mutually exclusive. The steps needed tomove toward a negotiated solution of the civil war would undermine the government coalition. Attemptsto revive the economy would hinge on finding a new source of external support, while any externalsupport is bound to have political overtones which would threaten the delicate balance of powerachieved after the elections." The al-Mahdi government was ousted in the summer of 1989.6. jamal (1988) argues that the IMF does not understand the functioning of the Somali economycharacterized by (a) a very large subsistence sector, and (b) the informal sector and the parallel marketwhich influences most urban transactions. Measures agreed between the Somali government and theIMF "are marginal to the functioning of the Somali economy. All in all the spectacle is one of the IMFtrying to impose the trappings of a free-market economy on Somalia whereas one already exists in allbut name" (p. 258). However, Jamal does not discuss what should be done to improve the Somalieconomy.


Table 4.10 Somalia: Chronology of Selected Policy Events, 1980-86

MonthlYear Event

February 1980 Macro policy agreement with IMF; implementation was incomplete.July 1981 Macro policy agreement with IMF; completed July 1982.July 1982 Macro policy agreement with IMF; completed January 1984.February 1985 Macro policy agreement with IMF; implementation was incomplete.March 1985 Debt service rescheduled by Paris Club.November 1985 Consultative Group of Aid Donors reviews economic policies.June 1986 Agricultural policy agreement with IDA.

construction of the large Bardhere Dam, aimed at exploiting the waters of the Juba River,was postponed pending the completion of a number of studies.

The Government of Somalia decided, for reasons which remain obscure, to abandon itsreform program in 1984.7 The country had suffered a major setback in the previous year as aresult of drought conditions and a Saudi Arabian ban on livestock imports from Somalia(exports of live animals accounted for 77 percent of Somalia's total exports and the volume ofthis item shrunk by 53 percent from 1982 to 1984). A large budget deficit triggered highinflation, and arrears on debt service started to accumulate.

In early 1985, the government adopted a major reform program including a substantialdevaluation and the establishment of a legal free market for foreign exchange for mostprivate transactions. A number of other moves toward liberalization also were made. Thegovernment negotiated an agricultural sector adjustment program with the IDA. It presentedits "National Development Strategy and Programme," including a revised three-year publicinvestment program for 1986 to 1988, to the Consultative Group in November 1985. Once againthe new public investment program was criticized for being too large. A foreign exchangeauction system was initiated in September 1986, but its operation encountered severedifficulties owing to its complete dependence on external aid. Many different exchange ratesapplicable to different types of transactions came into existence.

To conclude, the history of economic reforms in Somalia was far from straightforward.The government did not or could not sustain a coherent policy posture. Accordingly, Somalia'sclassification in Table 4.5 is different in various years. Somalia is included in the group ofcountries making a medium-level reform effort in 1981 to 1983 and once again in 1985 and in1986. In other years Somalia is classified as having low policy reform intensity.

Tanzania was substantially affected by the economic crisis in the late 1970s and early1980s because of large policy distortions and because of adverse, exogenous shocks ofrelatively low intensity (see Table 2.11). Yet not until the middle of the 1980s did the

7. According to Legum (1986, p. B 359), "Divided opinion within the Government led to a breach withthe IMF, which developed into something of a suspense drama. In March 1984, the Government hadsigned a letter of Intent accepting the terms of a new $183m IMF extended credit facility to run forthree years. However, at a meeting in April, the Council of Ministers decided by one vote to repudiatethe agreement, as its conditions were not acceptable. The major objection was said to be the IMFdemand for a 60 percent cut in the military budget. A further devaluation of the Somali shilling and cutsin government personnel were also demanded." The cabinet was reshuffled in June 1984. "The Ministerfor Information, Mohamad Omar Jays, explained that the motive for the change was to encourage unityand get rid of the 'confusion' prevailing during the last seven years-i.e., since the end of the OgadenWar, the Somali Government began its shift from its Russian alliance into the Western camp. Anexample of the 'confusion' was probably the rejection by the Cabinet in April of the terms for the IMFloan which they had previously all but accepted" (p. B351).


government mount a credible reform program, and this delay imposed an enormous cost on theTanzanian people and compounded the difficulties of securing an economic recovery. I willexplore the underlying causes of the initially muted policy response to the crisis bydescribing actions that the Tanzanian government took and by analyzing the protractedimpasse in negotiations that prevented agreements with international agencies.

A policy agreement with the IMF in 1980 was suspended after one quarter owing to theinfringement of the credit ceiling (see Table 4.11). The government formulated an exportrecovery program in 1981, supported by IDA, but this was mainly intended as a bridge to afull-pledged reform program. The government's National Economic Survival Program, 1981-82, had a similar provisional character. It proved to be ineffective (Wangwe 1987, p. 151).Meanwhile, the govemment appointed a special team of foreign and local experts, financedby IDA, to advise on the elements of a comprehensive policy package. A StructuralAdjustment Program, 1982-85, was promulgated by the government based on the work of thespecial team. Although a number of useful measures were implemented in 1982 and 1983under this rubric, the package did not pass the critical minimum test of a reform effortapplied by the IMF and the World Bank.8

Table 4.11 Tanzania: Chronology of Selected Policy Events, 1980-86

Month/Year Event

September 1980 Macro policy agreement with IMF; implementation was incomplete.April 1981 Export rehabilitation program agreed with IDA.July 1986 Consultative Group of Aid Donors reviews economic policy.August 1986 Macro policy agreement with IMF; completed February 1988.September 1986 Debt service rescheduled by Paris Club.November 1986 Multisector policy agreement with IDA.

According to Wangwe (1987), negotiations with the IMF were blocked for many yearsbecause the two sides could not agree either on the genesis of the crisis or on how to overcomeit.9 The IMF insisted at the outset on getting to "equilibrium" prices immediately, while theTanzanians did not even admit that any adjustment was required. Views inside the IDAoccupied the middle ground. IDA staff maintained that there was need for substantialadjustment in the exchange rate, the interest rate, and agricultural producer prices, but that

& Ajit Singh (1986) concludes that "many of the steps which the Tanzanian government has taken tocorrect the structural disequilibrium of the economy are in the same direction as those recommendedby the IMF. There are, however, two crucial differences: the first is the speed of adjustment, and thesecond concerns who bears the burden of adjustment" (p. 448). He argues that the IMF programimplied an explosive increase in the price of sembe, the basic food, from 2.5 shillings per kilogram to 18shillings.9. "Both sides held their beliefs very firmly and rigidly for some time in a rather dogmatic way"(Wangwe 1987, p. 153). President Nyerere said on New Year's Day 1981 that "Tanzania is not preparedto devalue its currency just because this is a traditional free market solution to everything andregardless of the merits of our position. It is not prepared to surrender its right to restrict imports bymeasures designed to ensure that we import quinine rather than cosmetics, or buses rather than carsfor the elite. My Government is not prepared to give up our national endeavor to provide primaryeducation for every child, basic medicines and some clean water for all our people. Cuts may have to bemade in our national expenditure but we will decide whether they fall on public services or privateexpenditure. Nor are we prepared to deal with inflation and shortages by relying only on monetarypolicy regardless of its relative effect on the poorest and less poor. Our price control machinery may notbe the most effective in the world but we will not abandon price control: we will only strive to make itmore efficient. And above all we shall continue with our endeavor to build a socialist society" (Payer1983, p. 799).


it was also important to relax nonprice constraints to production simultaneously. IDAvisualized a gradual adjustment process involving price and nonprice variables, a processthat needed to be supported by adequate amounts of external finance. It took a long time forthe IMF and the Tanzanian government to agree along the above lines; the position of bothparties shifted gradually toward the middle ground.1

The initial posture of the Tanzanian government reflected a strong skepticism regardingthe effectiveness of price policy instruments (Ndulu 1987, p. 20). In 1979, the governmentorganized in Arusha a seminar with academics and public officials from all over the worldto express joint disbelief in IMF recipes of the neoclassical type. Tanzania seemeddetermined to chalk out an alternative adjustment path that would avoid economiccontraction and the undermining of its "Basic Needs" philosophy. There was hope in Dar-es-Salaam that some of the donors (including possibly the Scandinavians and even the IDA)would support the Tanzanian initiative. As time went by Tanzanians began to recognize thatmore and more developing countries were adopting IMF recipes and that donors were notbehind Tanzania's alternative design. Furthermore, command structures within Tanzania(rationing, price controls, state monopoly of agricultural marketing, parastatals) were fallingprey to the spreading economic crisis and to parallel markets as well as to corruption. 11

Important changes in economic policy were made in 1984. Priorities shifted in favor ofagriculture. Primary responsibility for agricultural marketing was transferred fromparastatals to cooperatives. There was a partial liberalization of imports. The shilling wasdevalued by 35 percent, and the scope of price control was narrowed. The first significantdose of liberalization was administered under President Julius Nyerere, classified by Jacksonand Rosberg as a "prophetic ruler." Nyerere shifted his support from party colleagues of theold guard, who were wedded to the "command model," to technocratic advisors in thegovernment who were ready to experiment with a somewhat market-oriented policyframework (Hartmann 1988, p. 177).

Not until Hassan Mwinyi took over as president late in 1985 did Tanzania make adecisive switch. In August 1986, the government signed an agreement with the IMF thatincluded (a) significant adjustment of the exchange rate, (b) the target of moving toward anappropriate exchange rate by 1988, (c) an upward adjustment in the interest rate, (d) thegoal of securing a positive interest rate in 1988, (e) further reduction in the scope of pricecontrols, and (f) the objective of establishing full decontrol by 1988 except for twelve items.In November 1986, the government signed an agreement with the IDA abolishing allrestrictions on grain movements, allowing cooperatives to export directly, and improving theforeign exchange and trade regimes. The government adhered (as of late 1989) to theseagreements, which are part of its Economic Recovery Program.

Growth in real GDP in Tanzania slowed down to 1.2 percent per year during the 1980s.Expansion in total expenditure was curtailed sharply, largely by reducing investment. Theaverage annual rate of inflation accelerated from 14 percent during 1973 to 1979 to 31 percent

10. According to Green (1986, p. 41), IMF rigidity on devaluation, producer prices, and food subsidies"had a very high price. It has misdirected Tanzanian energy from phased solutions... to arguing againstmassive one off ones, and it has created emotional political symbols out of what had been and shouldhave remained largely economic managerial questions. Thereby it has slowed change in directions boththe Fund and (even for different reasons and on a phased basis) many Tanzanian analysts andeconomic decision-takers wanted, allowing imbalances to worsen and making changes when theyhappened too lumpy and too late, thus minimizing their real economy incentive impact andmaximizing their inflationary and human hardship potential."11. "The state control of the economy has been weakened and this has tended to water down the effectof whatever distributional policies were officially adopted. Official control of pricing and de juresubsidies of basic consumer goods and services have been undermined. This situation has reduced thepolitical constraints on change from the official status quo since such a change may in effect legalizewhat is actually taking place. Thus, politically it is easier to remove food subsidies when consumers infact pay much higher prices on parallel markets, and to introduce cost-sharing schemes to improve thequantity and quality of basic services when 'free' services are deteriorating for lack of funds" (Ndulu1987, p. 24).


from 1980 to 1986. External reserves were drawn down to dangerously low levels, and arrearsemerged in the servicing of external debt.

Policy responses in Zimbabwe were dictated by the arrival of independence in April 1980.After a period of relative isolation during Ian Smith's regime and protracted guerrillawarfare, Zimbabwe negotiated new constitutional arrangements at the Lancaster HouseConference. The radical socialism espoused by the Zimbabwe African National Union(ZANU) and Robert Mugabe during the liberation struggle was set aside in favor of areformist platform aimed at national unity (Gordon 1983; Mandaza 1986). The platform'smajor goals were to (a) revitalize the long neglected communal areas (formerly Tribal TrustLands) where the bulk of the black population farmed, (b) expand access to education andhealth services for the blacks, and (c) retain the confidence of the white farmers and thetransnational corporations that dominated the mining and industrial sectors. In allprobability, Mugabe would fall into the Jackson and Rosberg category of a "princely ruler,"although some observers would prefer to characterize Zimbabwe as a nascent democracy.

The outcome up to 1982 was a spectacular expansion in GDP, investment, and imports. Thisboom could not be sustained. A stabilization agreement with the IMF was worked out in 1983,but it was suspended in early 1984 (see Table 4.12). Nevertheless, the government succeededin restoring external balance through a comprehensive system of foreign exchangeallocations. In 1984, restrictions were intensified on the remittance of profits and dividends.A substantial real depreciation in the exchange rate was secured. According to Kadhani(1986, p. 113), Zimbabwe adopted the IMF stabilization philosophy without deploying itsspecific recipes and without adhering to the detailed targets it favored. The governmentmade little progress in tackling the budget deficit, however. The economy had not succeededin securing a satisfactory rate of export growth or in attracting foreign private investment.Today the overall GDP is still growing far too slowly, and well-paid jobs are not keepingpace with the rapidly growing labor force.

Table 4.12 Zimbabwe: Chronology of Selected Policy Events, 1981-86

Month/Year Event

April 1981 Macro policy agreement with IMF; completed April 1982.March 1981 Manufacturing rehabilitation policy agreement with IBRD and IDA.February 1983 Manufacturing exports promotion agreement with IBRD.March 1983 Macro policy agreement with IMF; implementation was incomplete.

ZANU obtained 77 percent of the popular vote in the 1985 elections, but the results weredisappointing from the standpoint of Mugabe's policy of reconciliation (Sithole 1986, p. 92).A majority of the whites voted for Ian Smith's Conservative Alliance of Zimbabwe ratherthan for the Independent Zimbabwe Group, which was more in tune with black rule. Inaddition, the people of Matabeleland registered a decisive preference for Joshua Nkomo'sZimbabwean People's Union (ZAPU). Later the two major parties (ZAPU and ZANU)merged under the leadership of Mugabe, thereby eliminating interparty politicalcompetition. The Constitution, however, had not been changed (as of late 1989) to establisha single-party state.

Economic policy is not much affected by the ethnic split between ZANU's supporters whoare Shona speaking and ZAPU's Ndebele-speaking constituency. It is, however, considerablyconstrained by the skewed distribution of assets and incomes between whites and blacks,which government has not altered for the sake of national unity. In this context, thegovernment feels politically compelled to retain administrative controls over the allocation


of foreign exchange and over the prices of wage goods. Through these controls, thegovernment hopes to influence the pattern of investment and production and to project to itssupporters a posture of being in charge. Liberalization, it is feared, may marginalize thegovernment's role either in reality or at least in the perceptions of many Zimbabweans.

After a long and tortuous discussion of trade policy, the government and the World Bankagreed in March 1981 to undertake a study of the government's import licensing and tariffpolicy in the context of a loan from the IBRD-IDA entitled the ManufacturingRehabilitation Imports Program. Much delayed, the study was not completed until thesummer of 1983 and resulted in a finding that an efficiency criterion should be used inallocating imports. It failed to address the question of how the government should liberalizethe import regime, although this was part of the understanding with the World Bank.Meanwhile, the government launched a short-term manufacturing export promotion programin the context of a loan from the World Bank. The focus of this program was on streamliningforeign exchange controls, providing an export subsidy to manufacturers, liberalizing exportcredit facilities, and maintaining a competitive exchange rate. The government hoped thatthis interim program would lead later on to a broad-based adjustment program based ondetailed economic and sectoral analysis.

Such a study program was not undertaken, however, for quite some time. The neglect ofmajor adjustment issues, including import liberalization, reflected the prevailing atmosphereof indecisiveness. There was, for example, considerable skepticism about liberalizingimports. It was feared that such a move would lead to deindustrialization, unmanageablepressures on the balance of payments caused by a flood of luxury imports, and enrichment ofaffluent whites at the expense of impoverished blacks. The World Bank urged thegovernment to undertake a systematic study aimed at a serious analysis of these questionsthat would also illuminate the implications of adopting different policy alternatives underthe generic rubric of import liberalization. It appeared in 1986 that the government wouldagree to such a study but on an extended timetable, thereby postponing significant policyaction for at least another year. It was not until July 1987 that the government announced adecision, in principle, to institute a process of trade liberalization. How and when thisdecision is to be implemented still remains vague.

There was very little willingness on the part of the Ethiopian authorities to negotiatepolicy agreements with the IMF and the IDA during the 1980-86 period. A number of policyreports were written by the IDA or jointly by staffs of the IDA and the government. Avariety of issues (notably agricultural prices, agricultural marketing, and the exchange rate)were discussed, but no agreement on them was reached. This protracted impasse reflectedbasic differences in orientation. The Ethiopian posture was influenced greatly by dramatichistorical events: the military coup in 1974 that ousted Haile Selassie, the far-reachingland reform, the widespread nationalization in 1975, the doctrinal commitment to scientificsocialism in 1976, the 1977 campaign of "red terror" against counterrevolutionaries and leftistintellectuals, the creation in 1979 of the Commission to Organize the Party of the WorkingPeople of Ethiopia (COPWE), and finally the establishment of the Workers' Party in 1984.Ethiopia became more of an orthodox socialist state with a vanguard party than almost anyother country in Africa. According to Schwab (1985), Mangistu Haile Mariam's leadershipwas of critical importance in guiding the revolution and it was derived from Leninist theory(p. 116). Magistu would fall into the Jackson and Rosberg category of a "prophetic ruler."

The main preoccupation of the Ethiopian leadership in the 1980s was to institutionalizeand legitimize the revolution, which had a distinctly military origin. Bloody battles hadbeen fought not only against Eritrian secessionists and Somali irredentists in the Ogaden butalso numerous other dissenters. Economic issues occupied a backseat. Staff of theinternational agencies had most of their contacts with economic ministries who played avery subordinate role in policy decisions. According to Schwab (1985, pp. 53, 54):

The government ministries in Ethiopia have traditionally been relatively powerlessand there is no evidence to show that the situation is otherwise at the present time


(1984). Their primary function is to carry out Dergue (the revolutionary committeeheaded by Mangistu) or COPWE edicts, and to deal with relatively innocuous day-to-day matters. They may, as in any bureaucracy, pace the carrying out of decisions orinfluence their tone and style, but they play almost no role in establishing policy orinfluencing ideology.

The Ethiopian economy weakened considerably during the 1980-86 period. Real growth inGDP fell below an annual average of 1 percent, causing a considerable decline in per capitaGDP. Total consumption continued to rise at a higher rate than did GDP, although asubstantial deceleration occurred compared with the 1973-79 period. The external pictureworsened considerably. The current account deficit rose from an average of 1 percent to 7percent of GDP. The very comfortable margin of external reserves during the 1970s shrank,while the average debt service ratio doubled. The Ethiopian leadership must solve theseformidable problems sooner or later, either with the collaboration of international agenciesor on its own.

Moderately Affected Countries

The three countries moderately affected by the economic crisis-Malawi, Mauritius, andKenya-scored relatively well in terms of economic reforms. Policy responses by Mauritiusand Malawi during the 1980s were very strong in comparison with what happened in therest of the sample. These are analyzed in some detail in Chapter 5. Kenya was quick andpersistent in adopting a series of policy packages, but it faced some obstacles inimplementing these new measures.

Kenya began the 1980s with an enviable reputation for political stability. Jomo Kenyattahad presided over the republic almost since independence. Although there was only onelegal political party, elections were fought vigorously and frequently resulted in the defeatof hitherto influential politicians. The Kikuyu tribe dominated the scene both politicallyand economically under Kenyatta. Upon his death, Daniel Arap Moi became president in apeaceful political transition noted for its rarity in Sub-Saharan Africa. Kenyatta and Moiwould be classified as "princely rulers" by Jackson and Rosberg.

Kenya also enjoyed a reputation for following sensible, conservative economic policies andfor securing high rates of economic growth over the entire postindependence period. It wasthe first country to use the IMF's Extended Fund Facility when it was established in themid-1970s. Relations between the government of Kenya and the IMF and the World Bankhad been very good over a long period. Not surprisingly, given this history, the governmentnegotiated a series of macro policy agreements with the IMF throughout the 1980-86 period(see Table 4.13). The government signed a structural adjustment policy agreement with the

Table 4.13 Kenya: Chronology of Selected Policy Events, 1978-86

Month/Year Event

November 1978 Macro policy agreement with IMF; completed August 1979.August 1979 Macro policy agreement with IMF; implementation was incomplete.March 1980 Structural adjustment policy agreement with IDA.October 1980 Macro policy agreement with IMF; implementation was incomplete.January 1982 Macro policy agreement with IMF; implementation was incomplete.July 1982 Structural adjustment policy agreement with IDA and IBRD.March 1983 Macro policy agreement with IMF; completed September 1984.February 1985 Macro policy agreement with IMF; completed February 1986.April 1986 Consultative Group of Aid Donors reviews economic policy.June 1986 Agricultural policy agreement with IDA.


World Bank on the very day that the Bank initiated the new type of policy-based lendingin March 1980 and a second such agreement was negotiated two years later. There was nofollow-up policy agreement with the World Bank or IDA until 1986, however.

Kenya's policy reform is classified into the medium intensity category in Table 4.5because the country had considerable difficulties in implementing its policy decisions. Anumber of the early agreements with the IMF had to be cancelled.12 The 1983 agreement wasthe first one in recent years to be implemented fully. Kenya's stance on fiscal and monetarypolicy, which historically had been conservative, seemingly became somewhat expansionaryin the late 1970s and early 1980s before resuming its traditional slant in the 1983-86 period.According to van der Hoeven and Vandermoortele (1987, p. 14), senior managers in alleconomic agencies pursued conservative, monetarist policies during the mid-1980s.

Implementation shortfalls characterized Kenya's experience with its two policyagreements with the World Bank. There was inadequate technical preparation in thecontext of some policy moves. The core econornic agencies of government did not have enoughcoordinating capacity to manage the full scope of the two policy packages. Governmentattention was distracted by the attempted coup in August 1982 and by the severe drought in1984. Furthermore, it appeared that a number of policy decisions were taken withoutreaching a critical minimum consensus either within government or vis-a-vis powerfulpressure groups.

One notable intent of agreements with the World Bank was to rationalize the system ofindustrial protection by abolishing quantitative restrictions on imports and graduallyreducing the average level and variability of tariffs. The government was committed toreducing the number of items subject to quantitative restrictions from 60 percent of the totalnumber of import items in 1982 to 12 percent by 1985. The tariff was to be revised by mid-1983and phased in during the next two years. Actual progress fell far short of these targets(Gulhati 1990). Import prohibitions were abolished in 1980, and tariffs were raised to amaximum of 100 percent on these items. In 1982, 20 percent of import items were liberalized,but acute foreign exchange difficulties led the government to reverse this move a few monthslater. There was a long delay in carrying out studies required for revising the tariff structure.The momentum of policy change was lost in this area partly because of opposition fromwithin the government.

Another controversial issue affecting Kenya's implementation of policy agreements withthe World Bank was the government's attempt to liberalize maize marketing by removingmovement restrictions on private agents and by redefining the role of the National Cerealsand Produce Board (NCPB) as a buyer and seller of last resort. A consultant's study in thisconnection was to be completed by March 1983 followed immediately by the formulation of agovernment action program. In the event, the study was completed in October 1983, the actionprogram was defined by December 1983, but implementation was held up for many months. Acomplicating factor was the 1984 drought. Furthermore, there was significant opposition frompolitically prominent large maize farmers who would have had to forgo substantial rentfrom the control system and to compete with Asian traders (Mosley 1986, p. 110).13 Progress

12. Of course, cancellations could be the result not only of implementation problems but also of designimperfections and technical snags. According to Killick (1983, p. 407), "While they are by no meanshostile to the basic task of the Fund, Kenyan officials dislike the way in which that role is played. Theycontemplate an application for an IMF credit with reluctance and regard negotiations with it asunnecessarily taxing."13. Mosley (1986) argued that in urging reform of maize marketing, the World Bank was "attempting afeat of political muscle which had defeated all liberalizing pressures from inside for over forty years' (p.112). He also expressed the view that there was no strong case for privatizing maize marketing sinceproducer prices were not depressed. He recognized, however, that the Bank wanted the price to varybetween surplus and deficit areas and that this was contrary to the long-cherished principle of pan-territorial pricing. He concludes, "But, with hindsight, it might have been politically wiser for the Bankto try and reform the NCPB gradually, by getting it to raise its average producer price a little and tomove gradually away from the principle of pan-territorial pricing, than to attack it frontally by trying topersuade it to abandon all powers save those of a trader of last resort. That is, the cost of getting a state


was made in early 1986 when the government allowed direct sales from farmers to millerswithout the intermediation of NCPB and trading by cooperatives. Some NCPB procurementcenters were closed, and private traders were licensed.

Kenya made a great deal of progress in restoring financial balances mainly by restrainingaggregate demand. The volume of investment declined by 3 percent per year during the 1980-86 periods, while consumption was brought to a standstill. The current account deficitaveraged 5 percent of GDP compared with 7 percent in the 1973-79 period. By 1986, thisdeficit had shrunk to only 1 percent. Substantial progress was also made in reducingdistortions in factor prices. The attempt to tackle structural issues was much less successfulsince it was difficult to mobilize enough domestic support for such reforms.14 The annualexpansion in real GDP declined from 4.8 percent in the 1973-79 period to only 2.9 percent inthe 1980-86 period. The volume of exports continued to decline throughout the latter period.Manufactured exports in real terms were 72 percent lower in 1982 than in 1974. Theyrecovered by only 12 percent between 1982 and 1985.

ConclusionThe policy responses summarized in Table 4.5 cannot be explained in terms of the rational-

actor model implicit in the views of applied welfare or development economists. Such amodel would have predicted a much more uniform response to policy distortions thanactually took place between 1980 and 1986. One could even have hypothesized a strongerreform effort (both in terms of intensity and continuity) in countries that were severelyaffected by either the shocks or policy distortions than in moderately affected countries. Butthe facts do not support such a hyphothesis. The evidence points to the concentration ofpolicy reforms in moderately affected countries such as Mauritius and Malawi. I have notdiscussed Botswana, but it, too, made a strong policy effort despite being a moderatelyaffected country.

There was no doubt that historical and political factors materially influenced the fate ofeconomic reforms. Deep-seated divisions and resulting civil strife affected the course ofevents in many countries. Pressure groups opposed to reforms demonstrated considerablecapacity to undermine the implementation of policy decisions in the Sudan and Kenya. Rigiddoctrinal commitment to statist policies in Ethiopia and pre-1985 Tanzania stood in the wayof meaningful responses to economic challenges.

The sample examined here does not provide much comfort to those who believe thateither democratic or authoritarian political systems facilitate economic reforms. The onlydemocratic country in the sample, Mauritius, carried out sustained and intensive reforms, butthis was also true of Malawi under an autocratic ruler. Two other autocracies, Zaire andSomalia, failed to generate strong policy responses, however. It was clear that capacity andwillingness to make economic reforms could not be linked directly to such systemnic features ofcountries. The record of "princely" rulers (in Uganda, Zambia, the Sudan, Zimbabwe, andKenya) was equally mixed and did not permit the drawing of useful inferences.

Economic reforms were much affected by external forces, such as developments ininternational markets of primary commodities, and by actions of foreign actors. The periodunder review saw the expanding influence of international agencies in the economic reformprocess in Africa. A very large number of policy agreements were made with the IMF onmacro issues, but many of these could not be implemented fully. The IBRD and IDA also

institution to simulate the operation of the market may be lower than the costs of privatizing it outright"(p. 116).14. According to Godfry (1987, p. 619), "The [World] Bank seems to have overreached itself in Kenya in1982. The dozen or so conditions attached to its second structural adjustment loan reached into almostevery corner of government policy, including many of the most controversial. Faced with such a list andin the midst of perhaps the worst economic crisis since independence, the then Finance Minister didwhat most Finance Ministers would do in similar circumstances: he signed on the dotted line and kepthis fingers crossed."


became part of the policy circle; they specialized in medium-term policy questions at theeconomywide or sectoral level. In Chapter 6 I will examine briefly the role of thesemultilateral agencies and the issue of policy conditionality.

5THREE CASE HISTORIES OF REFORM:

MAURITIUS, MALAWI, AND ZAMBIA

The premise of this chapter is that in-depth policy histories of Mauritius, Malawi, andZambia during the 1980-86 period will yield insights additional to those presented inChapter 4. The three cases lie along a spectrum stretching from an aborted effort at one end(Zambia), reforms with mixed results in the middle (Malawi), and a seemingly successfulprogram at the other end (Mauritius). The aim of this comparative analysis is to identifyhistorical, political, and economic factors that explain these outcomes. I begin by introducingthe three case study countries and providing a preview of my conclusions about each of them.A few selected economic and social indicators are presented in Table 5.1. The rest of thechapter consists of a two-staged discussion of several aspects of the economic policy process.Readers interested in the details of these findings or their documentation are invited to readthree previously published studies (Gulhati 1989a; Gulhati 1989b; and Gulhati and Nallariforthcoming).

Brief Profiles

Let me start by recalling key facts about Mauritius. Real GDP rose at a very rapid rate inthe early 1970s. From 1973 to 1975 there was an unprecedented increase in the internationalprice of sugar, the country's major export, followed by a sharp decline from 1975 through1979. The government proved unable to curtail aggregate demand during the late 1970s, and abig deficit developed in the external accounts. This was financed by a drawdown of foreignexchange reserves and hard borrowing from the Eurodollar market. Furthermore, exports ofmanufactured goods, which had grown at an annual rate of 31 percent between 1971 and 1975,slowed down very considerably. There was a similar setback in the expansion of tourism.Confronted by these difficulties, the government substantially changed its economic policies.These reforms turned out to be very successful but only after weather and terms of tradeshocks during the early 1980s had accentuated the difficulties faced by policymakers. By1986, however, budgetary and balance of payment pressures had abated, and manufacturedexports had resumed their earlier growth tempo.Mauritius's success is attributable to three major factors. First, the problems it tackled wererelatively small compared with those confronting most other countries in the twelve-countrysample. It is much easier to overcome a brief and mild setback, such as the one thatdeveloped in Mauritius at the end of the 1970s, than to climb out of a deep hole. Manyeconomic processes tend to be cumulative so that once a protracted and severe disequilibriumdevelops, the chances of further deterioration tend to increase significantly. Second, it isremarkable that even a mild disequilibrium precipitated a societal response in Mauritius.Three different governments, including a coalition of left wing parties with no previousexperience in running the economy, made and implemented policy decisions. It was a notableachievement to sustain the basic thrust of the reforms in spite of so many party andpersonnel changes. As I will explain later, the country's political culture facilitates thecarving out of a consensus even though interest groups are well organized and there isconsiderable conflict among them. Third, the sharp economic recovery might not have beenfeasible in the absence of the cushion provided by the relatively large European Economic

59


Table 5.1 Selected Economic and Social Indicators, 1970-80

Indicator Mauritius Zambia Malawi

(percentage of GDP)

Manufacturing value added 13 16 12Exportsa 48 41 27Importsa 63 45 40Gross domestic investment 28 21 31Current government revenueb 20 24 19Total government expenditure 28 32 28

(units as indicated)

GNP per capita 1980 (US$) 1200 600 190Population growth rate (% p.a.)C 1.5 3.3 3.1Poor (as percentage of total population)

Urban 12d 25 e 25fRural 12d n.a. 85g

Enrollment ratioPrimary school 106 96 59Secondary school 51 17 4Tertiary level 2 n.a. n.a.

Life expectancy at birth (years) 65 49 44

n.a. not available.a. Goods and nonfactor services.b. Excluding grants.c. 1970-80.d. 1979.e. 1978.f. 1977.g. 1980Source: World Bank data.

Community (EEC) sugar quota. Mauritius's geopolitical advantages and its foreign policyunderpined its economic reforms.

In Malawi real GDP also rose at a remarkably high rate during the 1970s. There was alarge expansion in estate agriculture, which generated a rapidly growing volume of tobaccoexports. Malawi's terms of trade peaked in 1977 and then deteriorated sharply. The ensuingloss equaled 9 percent of GDP. The government did not maintain a disciplined demandmanagement policy, and deficits in the budget and the balance of payments were financed byburdensome external debt as well as by domestic inflation. These financial disequilibriarevealed some weaknesses in the framework of economic policies at the sectoral level. Thegovernment took some time to respond coherently to these difficulties, but an intensive andsustained effort was made over much of the 1980-86 period. Although these efforts yieldedimpressive accomplishments, Malawi continued to face formidable financial and economicdifficulties at the end of 1986. Implementation of some of the agreed reforms was delayed ordisrupted by adverse developments such as a decline in the terms of trade and the disruptionof international transport routes through Mozambique.

Three Case Histories of Reform 61

The mixed results of Malawi's intensive and sustained reforms can be attributed not onlyto "bad luck" in the form of a hostile international economic environment during theimplementation period but also to weaknesses in the technical design of policy reforms andto problems in public administration. The Malawian policy process has both the strengthsand weaknesses of extremely centralized decisionmaking. The strength is that policy can bemade quickly, and there is little need to compromise with competing power centers. Theweaknesses are that adverse consequences of policies may not be reported, therebydiminishing the likelihood of mid-course corrections, and that the bracing competition ofideas is absent. Furthermore, although some progress has been made in building up localinstitutional and professional capacity for policy work, a great deal more is required todefine policy packages that are fully adapted to Malawian realities.

To complete the trilogy, let me review the stylized facts about Zambia. Its veryspecialized economy is dominated by copper. Historically, the economy had becomeaccustomed to a high level of imports financed by favorable copper prices. But 1974 saw amajor decline in the terms of trade, and the economy suffered a catastrophic loss equal to 23percent of GDP from 1975 to 1979. Problems were compounded by egregious policy failures inmanaging aggregate demand and in securing a sensible allocation of resources. No coherentpolicy program was mounted to deal with these issues. Instead, the country borrowed largeamounts on hard terms from foreign creditors. Failure to adjust left an onerous legacy ofoverindebtedness and rapid inflation. Sporadic policy moves to deal with the quicklydeteriorating situation were of no avail. Then in 1983, nine years after the sharp setback incopper prices, the government finally made a decisive change in its policy stance. It realizedthat it could not mount a comprehensive reform program owing to its very limited capacityto make or implement policy. Nevertheless, the government made intensive policy efforts atmacro and sectoral levels during the 1983-86 period, starting from a point at which theeconomy was suffering from acute disequilibrium. An overwhelming shortage of foreignexchange during the entire implementation period undermined the beneficial impact ofadopted policies. On May 1, 1987, President Kenneth Kaunda announced that Zambia wasabandoning the reforms which, he said, were disintegrating the social fabric of the country.

Four factors contributed to the Zambian impasse. First, the problems that Zambia tackledwhen reforms started in earnest in 1983 were very formidable compared with those in eitherMauritius or Malawi. Second, the policy process underlying the reforms was flawed. Thecapacity and willingness of the country to carry out reforms were much reduced by aweakened political leadership and administration. Furthermore, several pressure groups hadacquired a stake in the prereform policies, and they had considerable muscle to resistreforms that they regarded as a threat. Third, the policy packages adopted between 1983and 1986 suffered from a number of technical weaknesses. Fourth, Zambia's reforms were notadequately supported by foreign donors and creditors.

Extent of Disequilibrium

Zambia was classified in Table 2.11 as one of the countries "most severely affected" bythe economic crisis, in contrast with Malawi and Mauritius which were in the "moderatelyaffected" category. To back up these conclusions, I provided information about each of thesecountries regarding growth trends (Tables 2.1, 2.2, 2.3), the impact of shocks (Table 2.7), andindices of financial disequilibrium (Tables 2.4, 2.5). Furthermore, some features of the policyframework affecting the public sector, exchange rates, agriculture, and industry at the end ofthe 1970s were analyzed. Zambia's policies were described as highly distorted while thoseof the other two were only mildly affected. My emphasis in this section is on the historical


and political factors underlying economic policy changes during the postindependence period.These policy shifts contributed to the varying acuteness of disequilibria in the three cases.

Mauritius

Mauritius inherited a constitutional system along the lines of a Westminster-styleparliamentary democracy. Apart from the 1972-76 period during which an emergency wasdeclared and elections postponed, the country has adhered to the rules of this constitution.Many political parties represent a society deeply stratified along class, ethnic, and religiouslines. No single party has ever secured an absolute majority in parliament to form agovemment on its own. The compulsion to work together across party lines has been everpresent. Even within a single party, it has been necessary to build a consensus since all majorparties are loose agglomerations of diverse interests. These political parameters have put adistinctive stamp on the economic policy process. Prime Minister Seewoosagur Ramgoolamheaded four governments and developed a policy style that emphasized accommodation andreconciliation.

About 2 percent of the population are Franco-Mauritians who own large sugar interests,big commercial firms, and enterprises in the Export Processing Zone (EPZ). Initially, theywere linked politically to the Parti Mauricien Social et Democrate (PMSD). Creoles (ofmixed European and "colored" descent) are the second largest ethnic group. They are skilledartisans, dockers, and fishermen. Traditionally, they too were tied to the PMSD. TheChinese are a relatively small community (3 percent of the population), but they are a keycommercial and industrial group. At the outset, they too were allied to the PMSD.

Indians (Hindus and Muslins) are the largest ethnic group (68 percent of the population).A small number of Hindus are wealthy businessmen, professionals, and politicians. The vastmajority of Hindus are sugar estate and industrial workers and small planters.Traditionally, they were tied to the Mauritius Labour Party (MLP), but many crossed over toMovement Militant Mauricien (MMM), a radical party with links to major trade unions. TheMuslims (16 percent of the population) are mostly traders in urban areas. They wereassociated with the Comite D'Action Musulman (CAM) until 1971. Later some crossed over tothe MMM.

Franco-Mauritians had economic power, but at independence political power passed fromthem into the hands of Hindus, who had come to the island as indentured labor and whohad suffered a great deal during the colonial period. Such a schism could have producedradical, leftist policies, but it did not. The commitment to parliamentary democracyencouraged all parties to seek the middle ground. Mauritius's policy framework wasbasically a compromise. As noted in Chapter 3, the country ended up with an"accommodation strategy" (Rothchild and Curry 1978).

The labor constituency had to be satisfied with five policy planks. First, the govemmentcommitted itself to reducing unemployment, which averaged 20 percent in 1971. Thegovernment recognized that fiscal incentives and import protection provided to manufacturersproducing for the home market under the Development Certificate (DC) scheme had notyielded many new jobs. Attention turned, therefore, to export of manufactures, as in Jamaicaand Hong Kong. The Mauritian Export Processing Zone was established. Incentives in theform of tax holidays, exemption from import duties and from some aspects of the regulatoryregime, as well as preferential credit were provided to foreign and domestic investors whowholly specialized in exporting. Employment in the EPZ rose at an annual rate of 38 percentfrom 1971 to 1975, but then the rate of growth declined to 8.5 percent from 1976 to 1980. Thegovernment also adopted the policy of attracting "high-class" tourists from Europe and SouthAfrica. The number of tourists rose at an annual rate of 25 percent during the 1970-75 periodand at a much lower rate of 9 percent from 1975 to 1980. To create jobs for unemployed youththrough relief works, the Development Works Corporation (DWC) was established; in 1976,it had 14,000 workers on its payroll. Overstaffing became a widespread feature of


parastatals during the late 1970s, and compensation levels in the public sector tended to berelatively high.

The second policy plank consisted of labor laws to improve working conditions. TheIndustrial Relations Act of 1973 instituted the system of remuneration orders (ROs) by whichthe National Remuneration Board established minimum wages, conditions of employment,and specifications of jobs for various categories of workers in the private sector.Superimposed on ROs was the system of cost of living adjustments (COLA), recommended bya tripartite committee consisting of government, trade unions, and employer organizations.According to ILO (1986), ROs did not have a major impact on wages except in the sugarindustry and the EPZ. In the former, nonwage benefits were exceptionally high, and wageshad to be paid on a year-around basis even though there was not enough work during thelean season. These features could be the result of pressures of strong trade unions in the sugarindustry. The remuneration orders in the EPZ provided a substantially higher minimumwage for men than for women (30 percent lower). Over 80 percent of EPZ workers werewomen. Average real earnings for workers in all sectors of the economy were restrainedduring the 1971-74 period when a state of emergency was in operation. Antagonized by thegovernment's repressive measures, many workers voted for the radical opposition party(MMM) in 1976 after the emergency was terminated. The next coalition government (MLP andPMSD) secured a razor-thin majority in parliament and was forced to accept labor demandsfor COLAs and bonuses despite the sharp deterioration in the overall economic situation.

The third policy plank consisted of consumer subsidies and price controls on a number ofitems. The subsidy on rice and wheat rose from 32 percent of the landed cost of imports in1975 to 60 percent in 1979. It absorbed about 21 percent of overall government recurrentexpenditure. The fourth policy plank was a rapid expansion of government expenditures onsocial services, mainly health and education. These outlays rose at almost 25 percent peryear in real terms from 1973 to 1979.

The fifth policy plank was taxation of sugar exports (in addition to progressive taxationof personal income) to finance rising public outlays. The sugar tax on large estates (exportingmore than 3,000 tons) rose from 6 percent of the gross value of exports in 1972 to 24 percent in1979. Small sugar producers (exporting fewer than 20 tons) were exempted, and the rate ofduty varied with the tonnage exported. The sugar tax became controversial. The governmentviewed it as an instrument for redistributing the rent accruing to Mauritius on account of thesugar quota it obtained from the EEC in 1975 as part of the Lome Agreement. Since sugarsubject to this quota could be sold at a preferential price (much higher than the world freemarket price), Mauritius obtained an "EEC dividend" from 1977 to 1979 equal to 12 percent ofits GDP. The government argued that passing the preferential price on to affluent owners oflarge sugar estates in Mauritius did not make sense either from the standpoint of equity orefficiency. The quota was a limited one, and strong incentives for sugar production wouldlead to excess production, which would have to be sold at the low world price in the freemarket. Large planters argued that the sharp hike in the rate of the sugar tax and thefinancial burden of labor laws had hurt their profitability far too much.

The Mauritian problem in 1979 had three major dimensions. First, aggregate demand hadbeen mismanaged since 1973. The exchange rate had been allowed to appreciate from 1975 to1978, and wage levels had become excessive. Second, the government had a number of goals(economic growth with efficiency, rapid expansion of employment, and reduction ofinequality), but it had not articulated clearly the internal conflicts and trade-offs amongthem. Questionable policies resulted. There were elaborate labor laws, large wagedifferentials unrelated to productivity, and trade union pressures of varying intensity thatimpeded the efficient functioning of the labor market. The sugar export duty had becomeexcessive, and its rate structure induced a break-up of large estates into relatively inefficientsmall holdings. There was open-ended protection against imports given to D.C. enterprises inthe manufacturing sector, at the expense of Mauritian consumers. Third, the "politics ofrepression" from 1971 to 1975 was followed by the "politics of accommodation" from 1976 to1979. In both periods, but particularly in the second one, the political imperative of


governing a deeply stratified society according to the rules of parliamentary democracy hadreduced economic policy to the lowest common denominator. The government coalition in thelate 1970s had tended to react passively to events. It had a diminishing capacity to leadsociety out of its growing economic and financial troubles.

The distortions in policy in Mauritius, mentioned above, are hardly peculiar to thiscountry. They exist everywhere in some form or shape. What was distinctive aboutMauritius in the 1970s was the relative mildness of policy distortions. Aggregate demandhad been mismanaged but not to the extent of precipitating a disastrous financial crisis as inthe Sudan or Zambia or Zaire. Factor prices in Mauritius did get out of alignment, but theextent of disequilibrium was moderate. The sugar tax became excessive, but it was not solarge as to lead to a large decline in sugar exports as happened in Ghana with respect tococoa exports or in Tanzania with respect to all exports.

Malawi

The political system in Malawi is noninstitutional and therefore sharply contrasts thatin Mauritius. The paraphernalia of parliamentary democracy inherited by Malawi atindependence was soon abandoned. According to Jackson and Rosberg (1982), PresidentHastings Banda (with a life term) is an autocratic ruler. He has decisively influenced thepolicymaking process and the substance of economic policy. He is sensitive to the economicinterests of expatriate planters, indigenous estate owners (a group that the president helpedcreate), peasant farmers, civil servants, and the Asian trading community, but he makes surethat none of them gains an independent power base. He discouraged trade unions and cooptedtheir leaders into the Malawi Congress Party (MCP) or the administration. The Malawianpolicy circle is a small one.

During the 1970s, President Banda strongly emphasized economic development overterritorial or ideological goals. A very high priority was assigned to agriculture. Wheninternational sanctions were imposed against tobacco exports from white-ruled SouthernRhodesia (present-day Zimbabwe), the president seized the opportunity and promotedtobacco cultivation in Malawi. A forceful effort was made to promote tobacco estatesfinanced directly and indirectly by the Agricultural Development and MarketingCorporation (ADMARC). Press Holding Ltd. (PHL), a private firm wholly owned by thepresident, established two subsidiaries to operate a number of big tobacco estates. The outputof all tobacco estates rose at an annual rate of 20 percent from 1965 to 1972 and 16 percentfrom 1972 to 1979. The official aim was to develop estates and smallholders, but in practiceestate development occurred to a considerable extent at the expense of smallholders. Togenerate large financial surpluses, ADMARC paid prices to smallholder producers of tobacco,groundnuts, and cotton-prices far below the border prices at which it sold thesecommodities. Smallholders responded by diverting land from cash crops to cereals and bybecoming wage earners on estates. When tobacco prices declined in the late 1970s, manyestates and PHL found themselves in serious difficulty. About 10 percent of the estates wereclosed down. Commercial banks took over the management of financially distressed estates.PHL was highly leveraged, and its inability to service its debt to ADMARC and thecommercial banks precipitated economy-wide difficulties.

President Banda insisted on self-sufficiency in maize. ADMARC paid smallholders aprice above the export parity but below the import parity. The ADMARC price was uniformthroughout the country, thereby subsidizing farmers in the North who were far away frommajor markets. ADMARC's consumer price for maize did not cover full costs, and the loss wasfinanced by profits in the tobacco account. After the fall in world tobacco prices, such cross-subsidization was no longer feasible and the maize policy regime contributed to a seriousdeterioration in ADMARC's financial position.

Remember that Malawi's long-run record of economic growth was remarkably good,despite its land-locked geography and an underdeveloped institutional structure. The policyframework that had evolved in the postindependence period had a number of weaknesses (of


which I have mentioned two very important ones), but on the whole it was very welladapted to the country's circumstances. Pricing of foreign exchange and labor had very lowdistortions. President Banda had insisted on maintaining the efficiency of the civil serviceeven if this meant slowing down the process of indigenization. Government expenditure as ashare of GDP rose from 25 percent between 1972 and 1974 to 28 percent between 1978 and 1980,mainly on transport and smallholder agriculture. Government outlays on rural developmentprojects did not result in satisfactory increases in production owing to heavy implicittaxation by ADMARC of the cash crops mentioned earlier, smallholders' very low rate ofadoption of hybrid or improved maize, and the emergence of acute land pressures. Imbalancebetween government capital and current outlays had affected adversely the productivity ofthese expenditures. There was considerable room for improving the efficiency of governmentand parastatal operations, but these problems were not as acute as in many other countries inSub-Saharan Africa. Furthermore, the scope of the public sector in Malawi did not expand asmuch as in other countries. The government did not establish parastatals to producecommodities. It did not nationalize foreign investors. Like Mauritius, Malawi pursued arelatively conservative "accommodation strategy" (Rothchild and Curry 1978). The extent ofdisequilibrium in Malawi's economy at the end of the 1970s was relatively low.

Zambia

The multiparty constitutional structure, inherited at independence in 1964, wasmaintained until 1972, when UNIP became the only legal political party. The policy circleconsisted of cabinet and party colleagues, including key civil servants, who representedsectional interests (for example, those of the Copperbelt vis-a-vis other regions) andideological interests (socialistic and Moscow-oriented vis-a-vis Western business oriented).Econornic policy was determined much more by historical, ideological, and political factorsthan by technocratic economic analysis. A trend toward presidentialism started in the firstrepublic and became much stronger in the second. The policy circle diminished in size, andPresident Kaunda increasingly made decisions based only on consultations with his ownadvisors in the state house. Ministries were bypassed. People in key positions in thegovernment or party were frequently reshuffled. Jackson and Rosberg (1982) classify Zambiaas a "princely state" within the broader rubric of "personal rule."

Political independence was regarded as an instrument for helping Zambians to secureeconomic benefits and power. Large scale nationalizations of manufacturing concerns tookplace in 1968 followed by copper in 1969. The number of parastatals increased from 14 to 147during the 1970s. Zambia adopted a "transformation strategy" in the classification proposedby Rothchild and Curry (1978). Many public investments were started without properscreening. Examples are the second fertilizer plant, batteries, brickworks, sawmills, andautomobiles assembly. Many industrial investments yielded negative value added at worldprices. An important cause of low productivity was overstaffing. During the 1975-80 period,parastatal jobs increased twice as fast as did production, while private firms let go workersmore rapidly than the decline in their output.

Government expenditure expanded rapidly during the late 1960s and reached 41 percent ofGDP by 1970. Although the declared emphasis of government outlays was on ruraldevelopment and the provision of "Basic Needs" services, in reality there was a markedurban bias and a proclivity to respond to needs of affluent households. Schools and hospitalstended to be overdesigned and thereby costly in terms of capital and imports. A substantialpart of government outlays was on subsidies, particularly those on maize and fertilizer. Themaize subsidy (difference between landed cost of imports and the subsidized price, as aproportion of the landed cost) increased from 11 percent in 1967 to 57 percent in 1971 beforedeclining to 38 percent in 1980.

Despite a severe shortage of skills, the government adopted a policy of very rapidZambianization, thereby reducing sharply the standards of the civil service. During theFirst Republic, the objective was to maintain a meritocratic and politically neutral


government staff under the auspices of an independent public service commission. Exceptionsbegan to be made to these principles with increasing frequency, and according to Lungu (1983,p. 367), most key positions were exduded from the purview of the commission in the SecondRepublic.

Zambian wage levels, particularly those in the mining sector, were pitched at very highlevels. The unification of pay scales of black and white workers in 1966, as a result ofproposals rnade by the Brown Commission, was achieved by raising the wages of the blacks.This fulfilled a promise made to the Mineworkers' Union of Zambia (MUZ) during theliberation struggle. After world copper prices collapsed, the government tried to restrainwages and to reduce consumer subsidies. These policy moves were opposed by trade unions,and on some occasions the government had to back down.

Although President Kaunda's philosophy of humanism favored vigorous ruraldevelopment, very little was done about it. Production of export and food crops languished.Agriculture received only a very small share of total investment. The terms of trade of ruralareas deteriorated considerably as a result of unduly low agricultural producer prices fixedby government and a policy of import protection that raised the prices of goods bought byfarmers. The main preoccupation of the government was to ensure a regular supply of maizeat low prices for urban centers. Emphasis was given to large-scale commercial farmers andmedium to large-scale emergent farmers, while little was done to develop the potential ofthe very large number of traditional farmers. The National Agricultural Marketing Board(NAMBOARD) and provincial cooperative unions (PCUs) were regarded as instruments forachieving the politically important aim of a low-cost maize policy rather than as means forrural development. Informed observers believe that the emergence of a new urban middleclass (business groups, civil servants, and formal sector workers) prevented the realization ofKaunda's rural development goal. Traditional farmers scattered over the countryside andfragmented by language and ethnic barriers found it difficult to protect their collectiveinterest. Many migrated to the cities. The urban population rose from 24 percent of the totalin 1964 to 48 percent in 1980.

Financial and economic disequilibrium grew in magnitude as time went by, and very littlewas done to address the numerous policy and institutional distortions. It became increasinglydifficult to prevent a sharp fall in the volume of imports. Although the real effectiveexchange rate depreciated by 20 percent during the 1970s, the government had to rely moreand more on quantitative restrictions on imports and the licensing of foreign exchange. Theuse of these administrative instruments involved delay and significant misallocation ofresources. There was no clear, strong signal to the economy to economize on imports or todevelop new exports. Existing capacity could not be utilized owing to an economy-wideshortage of imported inputs. Even copper exports suffered for lack of imported spare parts.

The Policy Turnaround and Its ImpactAccording to Table 4.5, Zambia responded in the early 1980s to the economic crisis with

low policy reform intensity. Its effort to improve policies was far short of what wasrequired. The policy program gained somewhat in intensity in 1983 and got into too gear inthe following year. High intensity reform was pursued for three years (1984 to 1986) beforebeing abandoned. The contrast with Mauritius and Malawi was a sharp one; both countriesadopted forceful, high intensity reform programs throughout the 1981-86 period. The aim ofthis section is to describe briefly these reforms and to explore the underlying determinants ofthese very different policy responses.

Mauritius

It is remarkable that the relatively mild disequilibrium that emerged in Mauritius atthe end of the 1970s evoked a societal response. Prime Minister Seewoosagur Ramgoolam waspressured and persuaded to change course. The open political system generated feedback to


party and cabinet members that was strong enough to create collective concern about thedeteriorating economic situation. Efforts were made to mobilize additional concessionalfunds, but these did not yield much result. Even so, the government could have adhered tothe accommodating stance of fiscal and monetary policy by borrowing on exorbitant rates orby accumulating payment arrears, but it chose to take the politically unpopular route ofrestraining demand. The turnaround occurred when the finance minister, whose warnings hadbeen ignored during the late 1970s, began to gain influence with the spending ministries. Anagreement was negotiated with the IMF in October 1979, visualizing depreciation of therupee, raising of interest rates, and reduction of food subsidies and of real wages (see Table5.2). Many of these measures were carried out, but the agreement collapsed largely as aresult of new shocks-severe cyclones and the second major increase in oil prices. Anotherstabilization program in 1980-81 continued the new emphasis on austerity.

Table 5.2 Mauritius: Chronology of Selected Policy Events, 1978-85

Month/Year Event

February 1978 Macro policy agreement with IMF; implementation was incomplete.October 1979 Macro policy agreement with IMF; implementation was incomplete.September 1980 Macro policy agreement with IMF; completed September 1981.June 1981 First structural adjustment policy agreement with World Bank.December 1981 Macro policy agreement with IMF; completed December 1982.May 1983 Macro policy agreement with IMF; completed August 1984.June 1983 Consultative Group of Aid Donors reviewed economic policy.December 1983 Second structural adjustment policy agreement with World Bank.March 1985 Macro policy agreement with IMF; implementation was incomplete.September 1985 Sugar policy agreement with World Bank.

Unemployment had climbed to 20 percent in 1982, and the government's austerity policyhad taken its toll in terms of lower real wages and a decline in GDP. A general election washeld in June 1982 in which the ruling coalition of MLP and PMSD was totally defeated by anew left wing alliance of MMM and Party Socialiste Mauricien (PSM). The stresses ofstabilization probably influenced this outcome, but there were a number of other issues aswell. Paul Berenger, the new finance minister, took office in the middle of theimplementation period of an IMF standby. All of the agreed targets of this agreement wereimplemented, despite earlier criticisms of IMF prescriptions by MMM spokespersons. Whenconfronted by the responsibilities of office, Berenger apparently showed a large measure ofopen-mindedness. He was impressed by the force of the logic inherent in the IMF recipe andadopted it despite its unpopularity in many quarters. Another political crisis occurred inMarch 1983 when the leader of the PSM took issue with the government's ongoing austerityprogram. As a result the finance minister (a leader of the MMM) resigned. Another generalelection was held in August 1983 in which a new coalition-Mouvement Socialiste Militant(MSM), MLP, and PMSD-defeated the MMM. During 1979 to 1986, three differentgovernments carried out reforms, but the basic thrust of new economic policies was notaltered. Pragmatism triumphed over ideology. The national mood seemed to be to solveproblems in a workmanlike manner.


The first step in the reform process was to stabilize the economy, mainly by curtailingaggregate demand. Attempts by the World Bank to get the government to tackle key sectoralissues did not make much headway during 1979 to 1982. For example, the government did notappoint a sugar commission to define policy in this important sector until December 1982, andit was not until February 1985 that the government adopted a Sugar Action Program. Tounderstand this long delay one must recognize the extreme sensitivity of the sugar issue inthe Mauritian social and political context. The various interests involved had longmistrusted each other. The trade unions and the government viewed with suspicion statisticsput out by the sugar planters. Furthermore, it was essential to build a measure of agreementon how and when to tackle what continued to be divisive issues. Full agreement on how toreform the sugar sector was not reached by the sugar commission, and two of the membersissued a dissenting report. However, the measures finally adopted by the government werederived from the two sugar commission reports and a master plan prepared by the Chamberof Agriculture. The tax rate paid by large sugar estates was reduced from 24 percent to 17percent in March 1985. Measures were taken to improve the accountability of sugarcorporations. The ban on mill closures was lifted to take advantage of significant economiesof scale through consolidation of sugar processing capacity. The government started toestablish a network of farm service centers to provide smallholder sugar producers withinputs and extension services.

The government's delay in tackling another thorny issue-trade policy-is alsoinstructive. No serious analysis was undertaken until 1983 despite the persistent urging byinternational agencies. In September of that year the government decided not to impose anyadditional quantitative restrictions on imports as part of the policy agreement with theWorld Bank. Existing quantitative restrictions were replaced by tariffs in two steps in July1984 and January 1985. The new tariffs were substantially below the tariff equivalent ofquotas. The World Bank had expected the government to prepare an action plan for reducingover time the average level and variability of import duties. (The effective rates ofprotection in 1983 had averaged 89 percent for the manufacturing sector, ranging from a highof 824 percent for electrical machinery to a low of -24 percent for food). However, thegovernment was not ready to make the next move. Further import liberalization, it feared,would trigger a loss of external reserves, a fall in public revenues, and a decline in jobs inimport substitution industries. Not enough consultation and consensus building among affectedparties had taken place. The World Bank and the government did not arrive at anagreement in this area until January 1987.

Mauritius can be credited with many achievements. The government stabilized theeconomy. The current account of the balance of payment showed a surplus in 1986 comparedwith an average deficit of 11.6 percent between 1979 and 1982. The fiscal deficit fell fromnearly 12 percent in the 1979-82 period to 6 percent in the 1983-86 period. Perhaps the mostsuccessful reform was the restoration of the EPZ's international competitiveness through apackage combining a flexible exchange rate with wage restraint. Unemployment was reducedto 4 percent in 1987. Progress was made in improving the financial viability of sugar estates,in raising yields of smallholder sugar farmers, and in increasing food production. Theeconomy's supply response during recent years has been impressive. Many problems remain tobe solved, but it is clear that the country has made remarkable progress.

The economy's sharp recovery would have been less impressive if Mauritius did not haveaccess to the relatively large EEC sugar quota. During the reform period (1980 to 1986), EECsugar prices averaged 54 percent above the international free market price. This EECdividend amounted to $800 million. It enhanced Mauritius's import capacity by 19 percent.The country's foreign policy underpined its reform effort. Mauritius was having to make netresource transfers to its commercial creditors during the reform period. Donor assistance couldnot offset the negative resource transfer. In the absence of the EEC dividend, the economywould have encountered a serious foreign exchange constraint, which would havecircumscribed the expansion of the EPZ. The lesson here is that reforming countries have to


play all their cards well, not just the economic ones. Links between economic and foreignpolicy must be recognized.

Malawi

The initial response of the government to growing financial and economic difficulties atthe end of the 1970s was tentative and halting. Economic policy was a highly centralizedactivity in which President Banda played the predominant role. Civil servants werereluctant to carry "bad news" to the president. "Loyalists" among the bureaucracy tended todominate "technocratic" Malawian officials. The World Bank and the IMF intervened inthis situation by producing diagnostic studies and by offering quick-disbursing external fundsconditional on policy agreements. This posture of the international organizations galvanizedthe technocratic elements within Malawi to bring their policy proposals before PresidentBanda for decision. Agreement between Malawi and the two international organizations wasfacilitated by the absence of ideological differences and by the openness of the president totheir policy proposals. The scope of the ensuing reforms was more comprehensive than inmost SSA countries. The sequence of policy packages was to stabilize the economy and thenfollow up with supply-side measures focused on reducing price distortions and (later still)with institutional improvements (see Table 5.3). Implementation of these new policies wasensured by the top leader's steadfastness of purpose and by the lack of pressure groups withthe capacity to resist.

Table 5.3 Malawi: Chronology of Selected Policy Events, 1979-86

Month/Year Event

October 1979 Macro policy agreement with IMF; implementation was incomplete.April 1980 Macro policy agreement with IMF; implementation was incomplete.June 1981 First structural adjustment policy agreement with World Bank and IDA.August 1982 Macro policy agreement with IMF; completed August 1983.September 1982 Debt rescheduled by Paris Club.March 1983 Debt rescheduled by commercial banks.April 1983 Fertilizer policy agreement with World Bank.September 1983 Macro policy agreement with IMF; implementation went smoothly until

April 1986.October 1983 Debt rescheduled by Paris Club.December 1983 Second structural adjustment policy agreement with IDA.December 1985 Third structural adjustment policy agreement with IDA.March 1986 Consultative Group of Aid Donors reviews economic policy.

A major accomplishment was the restructuring of Press Holding Ltd., which played apivotal role in the economy. The matter was extremely sensitive because the president (assole owner) was personally involved in all the complicated steps required to reorganizePHL's finances and to alter its institutional arrangements. The reform took a long time todesign. Studies were initiated in 1981, and it was not until the end of 1983 that decisionswere taken and then implemented. By 1985, PHL was making a sizable profit. There weremany pitfalls along the way and much delay, but the outcome was reassuring.

An important component of the reform program was to "get the prices right" forsmallholder farmers by removing the bias against export crops. Very little progress wasmade initially, owing to a presidential decision to raise sharply ADMARC's maize producerprice in 1982 following the drought-related poor harvest. Later relative prices for export


crops did improve, and there was a corresponding production response. Experience showed,however, that "right prices" were difficult to define, that technical relations between pricesand production were uncertain, and that various experts disagreed about what was anadequate level of maize production and stocks. It also became clear that smallholder outputwas subject to severe land and credit constraints. Effective technical packages for raisingproductivity were not available. It was not realistic, under these circumstances, to expect asubstantial aggregate supply response to "getting prices right." You could have more maize ormore groundnuts, but you could not have both.

At the macro level, the government managed to reduce considerably the magnitude offinancial imbalance. The budget deficit declined from an average of 16 percent of GDP in1980 and 1981 to 7 percent in 1985 and 1986. The current account deficit on the balance ofpayments shrank from an average of 21 percent of GDP to 8 percent during the same period.These deficits were curtailed mostly by reducing the absolute level of investment and bytolerating a fall in the per capita consumption level. The effort was made to cut back thereal volume of government expenditure. Time and again this objective was frustrated. Macrofiscal targets and requirements at micro and sectoral levels conflicted. These tensions arosepartly because of unforeseen circumstances (for example, disruption of the transport routeacross Mozambique to the sea), partly because commitments to reduce unproductive outlayswavered (it was very difficult to reduce government outlays on buildings, includingresidential facilities for the president), and partly because of a mismatch between reformsat macro and lower levels. Several categories of public expenditures that were essential forutilizing effectively the existing stock of public capital (that is, recurrent outlays foreducation, health, transport, and agriculture) remained at suboptimal levels.

Malawi's financial and economic difficulties were not resolved by 1986 despite intensiveand more or less sustained efforts to carry out economic reforms. Two factors were responsiblefor this outcome. First, the unfavorable terms of trade and the disruption of transportthrough Mozambique were a major negative influence. There was a sharp contrast betweenMauritius with its "EEC dividend" and Malawi with its "bad luck." A major lesson ofMalawi's sobering experience is that expectations about economic policy should not beunrealistic. It is, of course, important to improve economic policies, but it is clear that goodpolicies in Malawi cannot neutralize the adverse impact of powerful geopoliticalcircumstances or very unfavorable terms of trade.

Second, while reforms in Malawi are pointed in the right direction, the technical designof its economic policies needs improvement. For example, obtaining a substantial supplyresponse from the agricultural sector will require more than price adjustments. Measures, suchas land taxes and improved land tenure, will have to be deployed to utilize the pool of idleland in the estate sector. These are sensitive issues for the ruling group of estate owners. Totackle them will require a combination of confidence as well as security on the part of civilservants and an enlightened political leadership. While some progress has been made ininstitutional development, a number of constraints persist. The system of publicadministration encourages reactive and risk-averse behavior. Policy battles have to befought by civil servants without a great deal of protection from ministers. Key personnel arerotated frequently. Some civil servants who acquired considerable policy experience duringthe early 1980s had to leave public service prematurely, thereby preventing the seasoning ofthe civil service.

Zambia

In the late 1970s and early 1980s the Zambian government tried to stabilize the economy,but its efforts were half-hearted. None of these policy moves came to grips with thefundamentals. Senior politicians and officials did not share a common view about what waswrong with the economy. The broad economic philosophy of these influential policy actorsdiffered considerably.


A decisive change in the general policy climate occurred in October 1983 following thegeneral and presidential elections. The mandate of Luke Mwananshiku, the new financeminister, was expanded to include the function of development planning. President Kaundamade a number of other changes, but he did not carry out a wholesale change in the cast ofpolicy actors in the central committee of UNIP, the cabinet, or the civil service. This provedto be a stimulating block in retrospect, as I show later on. He did, however, hold week-longmeetings in each provincial capital to prepare the public for new economic policies. Thegovernment realized that it could not design or implement a comprehensive reform package,and therefore settled on a piecemeal, sequential approach. Against the background of aseries of stabilization programs negotiated with the IMF, the government worked out policypackages with the World Bank on mining first, then agriculture, and finally manufacturing(see Table 5.4).

Table 5.4 Zambia: Chronology of Selected Policy Events, 1981-86

Month Year Event

May 1981 Macro policy agreement with IMF; implementation was incomplete.April 1983 Macro policy agreement with IMF; implementation was incomplete.May 1983 Debt rescheduled by Paris Club.March 1984 Copper and export diversification policy agreement with World Bank.May 1984 Consultative Group of Aid Donors reviewed economic policies.July 1984 Macro policy agreement with IMF; implementation was incomplete.July 1984 Debt rescheduled by Paris Club.December 1984 Debt rescheduled by commercial banks.January 1985 Agricultural policy agreement with World Bank.June 1985 Consultative Group of Aid Donors reviewed economic policy.October 1985 Government announced foreign exchange auction.October 1985 Industrial policy agreement with IDA.December 1985 Consultative Group of Aid Donors reviewed economic policy.February 1986 Macro policy agreement with IMF.March 1986 Debt rescheduled by Paris Club.December 1986 Consultative Group of Aid Donors reviewed economic policy.

The main thrust of three macro policy agreements with the IMF was to stabilizeZambia's economy and regain financial balance very rapidly. For example, their aim was toreduce the balance of payment deficit from 12 percent of GDP to 6 percent between 1980 and1982 by cutting back on the volume of imports. No major increase in exports was feasible inthe short run. The budget deficit as a percentage of GDP was to be halved during the 1980-82period by reducing the volume of government expenditures by about 18 percent. Later, in 1984,an attempt was made to reduce external payment arrears by $100 million; this was a timewhen the real economy (including the copper mines) was heavily constrained by theunavailability of imported inputs.

Zambia could not achieve the targets spelled out in the macro policy agreements withthe IMF. The unrealistic goals that repeatedly were set tended to destroy the credibility ofthe policy process. Furthermore, assumptions about the behavior of international copperprices turned out to be erroneous on many occasions. They were expected, for example, todecline by about 6 percent in 1981, but instead they fell by 20 percent. Not even one of thestabilization programs succeeded. There was a long period separating the breakdown of oneagreement and the putting into place of the next one, largely due to disagreements onexchange rate policy.


The emphasis of the policy package on regaining financial balance in short orderrelegated the goal of growth in GDP to a back seat. The macro framework did not visualizeimprovements in real per capita GDP or restoration of investment to normal levels in thenear term. The objective of poverty alleviation was also given short shrift. These prioritiesdidn't make sense in the context of an economy that had suffered a protracted decline; thepopulation was growing rapidly, and competing pressure groups were greatly displeasedabout income distribution. Consequently, it would have been very difficult to securemeaningful domestic consensus to support such a reform effort.

The move to a market-determined exchange rate in October 1985 was a bold one, given thelack of experience with auctions and the skepticism among policy actors. The exchange ratejumped from 2.20 kwacha to the U.S. dollar before the auction to 7.37 kwacha in June 1986,15 kwacha in December 1986, and a peak of 21 kwacha before the auction was abandoned inMay 1987. The move to liberalize credit and foreign exchange (including the completeelimination of quantitative restrictions on imports) was made in the middle of an acutefinancial crisis. The new package generated some positive benefits: the disappearance of theblack market, a sharp reduction in corruption, and a new spirit among many economic agentsto raise the efficiency of their operations. However, liberalization at the macro levelcontrasted sharply with controls and rigidities at sectoral and micro levels (for example, themaize subsidy, controlled agricultural producer prices, the reluctance of the public sector tolay off redundant labor). Furthermore, it could be argued that liberalization of credit andforeign exchange markets should have come after efforts to bring the budget under control.Budgetary pressures intensified in 1986, and monetary policy acquired an accommodatingstance. Very low reserves, unpredictable capital inflows, and the overhang of externalpayment arrears made it very difficult to feed the auction with a stable source of foreignexchange. Some economic agents doubted that the auction would survive. Indeed, many keyofficials were perceived by economic agents to be working behind the scenes to resurrectcontrols.

Instead of the auction, the government could have decided to engineer a substantial realdepreciation in the official exchange rate over a specified time and a correspondingtimetable for phasing out import restrictions. If the initial target for real depreciationproved to be insufficient, it could be followed up by another round (Gulhati, Bose, andAtukorala 1986). This alternative would have relieved pressure on the overburdened controlsystem without the appearance that the government had abdicated its responsibility ofconserving scarce resources. It would also have allowed time for economic agents to adjust tochanges in the exchange and trade regime.

A critical aspect of the Zambian problem was that copper, almost the only source offoreign exchange, had encountered diminishing returns. The grade of ore reserves hadsteadily declined, and the cost of mining ore had risen. The current level of production couldbe maintained only for about seventeen years. A key aim of the reform was to rehabilitatethe copper industry, even though it had a limited economic life. Zambia ConsolidatedCopper Mines (ZCCM) drew up a comprehensive and operational production and investmentprogram that was approved by the government and the World Bank. Three mines wereclosed, and 4,000 workers were laid off in 1986. This retrenchment was on a much smallerscale than had been programmed, owing to resistance from the Mineworkers' Union ofZambia, but nonetheless an effort had been made to run the mines on a more economic basis.Agreement on the fiscal regime for ZCCM was delayed. The government found it difficult toresist the practice of extracting maximum revenue, thereby undermining the financialviability of ZCCM's rehabilitation program.

To secure a quick supply response from the agricultural sector, the government decided totackle problems faced by commercial and emergent farmers. These relatively affluentfarmers had developed land, management skills, tested technological packages, andphysical capital equipment. The production potential of these farms could be improved bysolving pricing and marketing bottlenecks and by injecting imports of fertilizer, spare parts,and other inputs. The government also wished to improve the condition of traditional


farmers but took the view that this would take much longer. The principle of economicpricing was adopted in 1982, and nominal producer prices were raised. However, average netincentives for major crops did not improve very much, if account is taken of exchange ratedepreciation as well as the remaining overvaluation of the kwacha (Jansen 1988).Responsibility for primary marketing of crops and agricultural inputs and for intraprovincialtrade in these items was transferred from NAMBOARD to provincial cooperative units. Thismove was not well prepared and marketing problems persisted.

To reduce the budgetary subsidy, the government raised the consumer price of maize from14 kwachas in 1980 to 54 kwachas in 1985. However, the establishment of the foreignexchange auction in October 1985 increased once again the kwacha cost of this subsidy,thereby greatly accentuating the pressure on the budget. The sharp depreciation of thekwacha had increased the cost of living very steeply. The original target for phasing outthe maize subsidy was postponed in mid-1986 on the grounds that adherence to the earliertimetable would have caused severe hardship. Nevertheless, the government moved aheadof the revised schedule when it decontrolled the retail price of breakfast meal (high-gradeflour) in December 1986 and doubled the price that millers had to pay for maize toNAMBOARD. The intention was to subsidize millers so that the retail price of lower graderoller meal could be maintained unchanged. Unfortunately, millers responded to thesegovernment moves by doubling the retail price of breakfast meal and by stopping theproduction of roller meal. They claimed that arrangements about how subsidies would bepaid to them for producing roller meal lacked specifics. The sharp increase in the consumerprice of breakfast meal and shortages of roller meal precipitated widespread riots.

The restructuring of agriculture raised very difficult technical issues. There wasconsiderable tension between retaining controlled producer prices and instituting a foreignexchange auction. Equally complex was timing the removal of the consumer maize subsidy,the reintroduction of spatially differentiated producer pricing, and the reorganization ofagricultural marketing.

The design of many of these policies was worked out largely outside Zambia by the IMF,the World Bank, and foreign consultants. A small group of Zambian officials and a fewZambian politicians spearheaded the reform based on a genuine understanding of its meritsand risks. These "reform-mongers" faced an acute scarcity of professional analysts and ofrelevant information, which added to the riskiness of the program. Most influential civilservants and politicians had very little understanding of the issues. Some segments of theZambian policy circle were hostile to the reforms for ideological reasons or because they hada vested interest in the status quo. The reformers got only intermittent support from theirbosses in their clash with influential "dissenters." Since the new policies failed to produceeasily recognizable signs of economic recovery, the reformers became increasingly vulnerable.Technicians could always claim that the economic situation would have been much worse inthe absence of reforms, but such arguments did not prove to be politically credible. Dissentersclaimed that the IMF-World Bank medicine being prescribed contradicted the socialistic andwelfare elements of Zambia's humanist ideology. They also pointed to the inequitabledistribution of the burden of adjustment. Some dissenters felt that Zambia (together withother African countries) was being exploited by the West. There was also the sentiment thatlocal elite groups had "hijacked" the Zambian state.

Finally, lack of adequate external support (in terms of amounts and timing) substantiallyundermining the positive impact of new policies. The World Bank and the IMF strenuouslytried to make their own funds available to Zambia between 1983 and 1986 and to mobilizeaid and debt relief from others to support the intensive reform process. Nevertheless, theresource transfer from the World Bank to Zambia was negative in 1983 and 1984; it becamepositive only in 1985 and 1986. In the case of the IMF, net resource transfers were positive inthe first two years and then turned negative. These outcomes reflected the policy,procedural, and resource constraints facing both organizations. The Bank Group convened fourmeetings of the Consultative Group of Aid Donors during this period. However, bilateralODA commitments were lower in 1983 to 1985 than in 1980 and 1981. They did increase


sharply in 1986. The net resource flow on account of suppliers' credits was negative and thatfrom commercial banks declined markedly. Altogether, the average net resource transfer fromall sources declined from $557 million in 1980 and 1981 to $300 million in the 1983-85 periodand it was $428 million in 1986.

The Paris Club rescheduled Zambia's official debt thrice, but altered debt servicepayments (on the basis of a grace period of five years, maturity of ten years, and interestrates linked to LIBOR) could not be honored. Arrears emerged quickly after eachrescheduling operation. Decisions made by the Paris Club did not necessarily reflect theconclusions of the consultative group or the financing plan proposed by the World Bank andthe IMF. The impasse that developed in mid-1984 accentuated Zambia's foreign exchangedifficulties (Jaycox et al. 1986, p. 6). The London Club rescheduling of commercial debt wasbased on an interest rate that was 2 1/4 percent above LIBOR-among the highest marginson recent reschedulings. Zambia could not honor these payments either. Preferred creditors,including the IMF and the World Bank Group, accounted for 45 percent of the total scheduleddebt service. Zambia could not keep up contractual debt service payments to these creditorsas well.

Aid and debt relief in support of Zambia's intensive reforms turned out to be too little, toolate. The international machinery was too fragmented to provide the necessary assurance toZambian leaders who had exposed themselves to risky economic reforms. There were toomany foreign actors pursuing diverse objectives to produce a coherent response (Gulhati andNallari 1988).

Conclusion

The three case studies of Mauritius, Malawi, and Zambia illustrate the varying role ofhistory in conditioning policy postures. For example, the racial divide between black andwhite in Zambia during the colonial period left an indelible mark on the entire economicpolicy framework in the period after independence. However, in Malawi, which had asimilar history, economic policy was less strongly affected by the race question. PresidentBanda emphasized "efficiency," eschewed nationalization of foreign investors, and avoidedprecipitate indigenization. In Mauritius the history of indentured labor affected politics andeconomic policy but subject to the "rules of the game" of a Westminster-style constitution.

The case histories also show how leaders and the body politic interact in the course ofdesigning and implementing economic policies. The singular role played by Banda in makingeconomic policy in Malawi, and in presiding over its implementation, is one type of modelwith its own merits and drawbacks. Ramgoolam in Mauritius also exercised a pervasiveimpact on economic policy but in a completely different way. He became the main instrumentof accommodation and reconciliation between strongly entrenched pressure groups and manypolitical parties. He had views about the substance of economic policies, but his maincommitment was to parliamentary democracy. Kaunda's leadership in Zambia does not fiteither the Banda or Ramgoolam models. He transformed a multiparty democracy into asingle-party state, thereby abandoning the constitution of the First Republic. He championedrural development at the doctrinal level but countenanced markedly pro-urban policies in theface of pressures exerted by a new constellation of interest groups. During the 1980s, he ledZambia belatedly into a reform program that was largely designed abroad, that wastechnically flawed, and that had insufficient political backing at home.

Finally, the case histories bring out the riskiness of reforms as well as the role ofexogenous shocks in determining their outcome. It would seem that Zambian reforms were agamble in view of the acuteness of the disequilibria, the lack of a reliable theory fromwhich to derive the technical design of specific policy measures, and the presence of well-organized pressure groups with a perceived stake in earlier policies. This gamble was lostpartly because adverse exogenous shocks occurred during the period of implementation andpartly because reforms were politically ill prepared. Reform efforts in Malawi also were


exposed to the special risks pertaining to land-locked countries. Furthermore, the technicaldesign of agricultural policies turned out to be flawed, although these weaknesses couldhave been remedied through a more systematic policy preparation effort. Mauritius, bycontrast, managed to secure a successful outcome because the problems that had to be solvedwere moderate in relation to the country's policymaking and policy implementationcapacity. Furthermore, the positive impact of the "EEC dividend" helped ameliorate whatcould otherwise have become a chronic foreign exchange bottleneck.

6FUTURE POLICY REFORM

During the relatively brief period since they gained their independence, the dozencountries in the sample have seen a great deal of change in their economic policies. The firstround of change occurred in the 1960s and early 1970s and its main thrust was governed bypolitical impulses generated by the liberation struggle, by the evolution towardnoninstitutional personal rule in most countries, and by the state-centered developmentphilosophy of the time. The second round of policy change began in 1980 and is stillcontinuing. After reviewing the first round, I focus on the period 1980 to 1986 to analyze thesecond round.

To a large extent, the policy change during the 1980s was a response to growing economicand financial difficulties. The reforms reflected the growing skepticism regarding the role ofthe state and the emphasis on market-oriented policies that had emerged almost as a globalphenomenon. These ideas were not indigenous to Sub-Saharan Africa; they came to thisregion through the intermediation of the IMF and the World Bank.

Policy change during the 1980-86 period included some successes and a large number ofcases in which results were mixed or distinctly negative. Particularly in Africa, policyreforms are a risky endeavor. Reforming governments or foreign policy actors can easilyderail them. They can also fail because of technical weaknesses in the design of policypackages or as a result of large, negative exogenous shocks. In this chapter I discuss measuresthat low-income countries and others can take to enhance the probability of successful reform.

Although economic policy reform can improve economic outcomes, it is not the only, oreven the decisive, variable. Its role should not be exaggerated. Much will depend on thesettlement of political disputes such as the one causing havoc in Southern Africa or the onedividing Southern and Northern Sudan. Furthermore, a major determinant of the economicfuture of Africa is the speed of technological progress, particularly in agriculture. Indeed,technical improvements can play a crucial role in resolving internal political conflict (Bates1988).1

Improving the Capacity and Willingnessof a Country to Undertake Reforms

In many countries in Sub-Saharan Africa, economic recovery requires a change in "societalarchitecture" (Dror 1986). Policy improvements of medium or low intensity will not suffice;they must be "transforming" or "reorganizing," in the terms of the Rothchild and Curry(1978) classification (see Chapter 3). Deep-seated realignment of economic, social, andpolitical forces will be required. The "development coalition," which is said by Bates (1988)to be underpinning the existing edifice of policy, will have to be altered. Key vestedinterests may loose rent, power, and status during the transition. In Africa, the transitionmay last for many years since unfavorable initial conditions (an aspect of underdevelopment)can slow down the impact of policy reforms. The immediate beneficiaries of policy change

1. Bates (1988, p. 128) states, "African governments want lower food prices. African farmers wantprofits. Under present conditions, the result is a deadlock, as farmers withdraw from markets renderedunprofitable as a consequence of government pricing policies. It is, of course, possible for the interestsof both parties to be served. Prices could decrease and profits increase were the costs of farming todecline. Investments in technical change therefore represent a way of 'decoupling' the central politicalconflict between farmers and the state in Africa."

77


are likely to be those with little political muscle (for example, peasant farmers).2 Forgovernments initiating reform, the imbalance in the political salience of losers and gainersand the protractedness of the policy packages' gestation period involve many dangers.

Past experience does not permit an optimistic assessment of the capacity and willingnessof SSA governments to reform their economies. Some scholars (Hyden 1988 is an example)have argued that for a nation to sustain far-reaching economic changes it must transform itspolitical system. It is difficult, however, to find any precise definition of the nature of therequired political transformation. History does not confirm a strong causal link betweenability and willingness to undertake economic reforms and any particular political systemsuch as democracy or authoritarianism (Haggard and Kaufman 1989, p. 233). According tothese two authors, "strong" authoritarian regimes insulate technocrats who design andadminister economic reforms from societal pressures, thereby enhancing their capacity tocarry out such changes; "weak" authoritarian regimes give less latitude to technocraticreformers who are subordinated to political elites and their pursuit of patron-clientrelations. 3 Furthermore, these authors maintain that "consultative" democracies are betterat economic reforms than "plebiscitary" democracies or even authoritarian states. BothMauritius and Botswana are examples of consultative democracies. Malawi would be anexample of a strong authoritarian regime and the Sudan of a weak one.

It follows that African governments can respond to a substantial extent to the challenge ofthe time without first becoming consultative democracies. The quality of leadership countsfor a lot in the context of authoritarian personal rule. Among the many inhibiting factorsthat hindered past reforms, the crucial one was the leadership variable. An authoritarianruler can make a decisive difference in determining the climate for economic reform. Hisstrategic vision or lack of it can make all the difference. The length of his time horizon isvital. Also important are the role he expects the civil service to play in the policy process,the extent to which he relies on patronage, the extent to which he is corrupt or toleratescorruption, and the ways in which he secures compliance with his reform agenda. Theprobability of intensive, sustained, economic policy change strongly depends upon theattitude of the ruler on these questions.4

What is the prospect that African rulers will change their approach to governing?Scholarly opinion is divided. Hodd (1987) strikes a hopeful note by emphasizing the impactof new ideas disseminated through Western universities and the international agencies toAfrican experts and through them to their leaders. He hypothesizes that rulers, having

2 Nelson et al. (1989, p. 8) state, "There is no one to one correlation between economic gains andpolitical support; much depends on how the gainers fit into existing political institutions and processes."3. According to Haggard and Kaufman (1989, p. 233), the typical "strong" authoritarian regime wouldhave the following features: continuity in leadership and/or relatively clear rules governing succession;a political structure that insulates technocrats and economic decision-makers from societal pressures,as well as from the demands of political elites themselves; an economic policy machinery with aminimum of capture by social groups; "corporatist" organization of interests through state-sanctionedand controlled associations; a military policy, or domestic intelligence network, capable of penetratingstrategic social institutions and deploying violence where "necessary."

A "weak" authoritarian state may share many of the formal characteristics of a strong one, but itwill also have these characteristics: frequent changes in leadership through "palace coups" or factionalrivalry within the ruling political elites; a low degree of insulation for technocrats from the politicaldemands of powerful social groups and the executive itself; a dualistic decision-making structure inwhich technocrats control only a limited range of policy instruments and compete with political eliteswho deploy other public resources for both political and private purposes; extensive networks of patron-client, personalistic, and familial relations within the formal government structure, sustained bycorruption, rent-granting, nepotism, and the discretionary allocation of governmental resources.4. Bates (1988, p. 125) argues that "governments in Africa may have learned from past mistakes. Butthe changes in their policy preferences may not result in changes in their policy choices. The situationwithin which they seek to retain power requires that to secure change they must alter basic beliefs andvalues, enlighten other major groups as to where their interests lie, and orchestrate reciprocaladjustments among them. Economic reconstruction poses daunting challenges to political leaders, andwe can deservedly be skeptical of the prospects of success."

Future Policy Reform 79

acquired personal riches, are now keen to take measures to improve the economic conditionsof their people. Cartwright (1983) points out that African leaders' rising concern for thepublic welfare may originate from their sense of insecurity. Protracted economic decline inmany SSA countries has meant a loss of general public support, and it has enhanced therulers' fear of losing power.5 Arguing along the same lines, Fearon (1988) has suggested thatrulers have a strong interest in preventing the collapse of the modern economy and of statemachinery. When they are threatened with such complete disintegration, rulers tend toreconsider their mode of operation.

My own view is that some combination of adversity, self-learning, and mutual learning(for example, the demonstration effect of reforms in nearby countries) can trigger changes inan authoritarian ruler's orientation that will then permit significant economic reforms tooccur. A good example is Malawi's President H. Kamuzu Banda. Another illustration isMadagascar's President Didier Ratsiraka, who took some initial steps toward reform in 1982but did not adopt a new, unambiguous policy stance until 1987 or 1988. Alternatively, anation may gain a new orientation toward economic reform with the succession in office of anew leader. This occurred in Ghana with the coming to power of Jerry Rawlings. Inauthoritarian regimes, particularly autocratic ones, the leader is much more likely toexercise the initiative for reform in a top-down policy process. The probabilities of a bottom-up approach tend to be higher in regimes that permit more participation-"princely" rule,plesbiscitary democracy, and consultative democracy. What happened in Mauritius at theend of the 1970s exemplified bottom-up policy change in a consultative democracy.

Even in the most open, pluralistic societies, however, policy reform will require majorinputs from the highest political authorities in the different branches of government.Interest groups of various kinds tend to have a narrow perspective, whereas policy reformfrequently requires a broad one. Furthermore, a crisis situation can promote a shortening ofthe time horizon of some special interest groups. By contrast, policy reforms frequentlyrequire a long-term perspective and an ability to see the economy as a positive-sum game.The task of leaders is to broaden the perspective and lengthen the time horizon withinwhich various groups assess costs and benefits of proposed changes. Leaders are more inclinedto undertake this task in response to a crisis than in "business as usual" situations. Hampson(1986, p. 23) contends, for example, that although micropolitical motives dominate decisionsin normal times, adversity tends to enhance the priority attached to macropolitical values.

Once the top leadership has become reform-minded, supportive changes in core economicministries (Finance, Planning, Central Bank) and major sectoral ministries can take place.These will take the form of changes in top-level personnel, new definition of responsibilities,changed incentive systems, and altered procedures required to initiate and sustain the processof economic reform. In recent years many SSA countries have attempted to reorganize theircivil service or parts of the government machine, but these efforts have been frustrated bythe top leaders' lack of sustained political commitment. It is this experience that leads meto conclude that a change in the orientation of the highest political authorities is aprerequisite for successful administrative reform.

Capacity for policy work can be enhanced by expanding the supply of indigenousprofessional expertise. In many African countries the number of university graduates hasalready increased, but shortages of high-quality experienced experts persist. To aconsiderable extent, this is the result of weaknesses in mobilizing and using available talent.These weaknesses reflect the sociopolitical context in which little emphasis is given totechnical inputs into policymaking. Once new signals are broadcast by the top leadershipacknowledging the need for high-quality professional work, it should be relatively easy to

5. Cartwright (1983, p. 292, 293) states, "Prosperity by itself did not guarantee survival; but its absenceseems to have made survival very difficult.... Mazrui's assertion that Africans seem to want leaders whomake a conspicuous display of wealth ... applies ... only to those leaders who have at the same timeprovided material benefits to their people."


organize the required changes in supply.6 This will involve improvements in incentives, on-the-job training, and management of professional cadres. Attempts to make theseimprovements without strong support from the top of the government pyramid are likely tobe unproductive.

Improved policy work requires not only some restructuring of the government machine butalso effective staff work preparatory to the building of new coalitions for reform. Top civilservice officials and senior advisors must learn how to deal with various layers of thebureaucracy, pressure groups, and government as well as party politicians. The quality ofthese interactions depends on a host of factors including mutual confidence, a sense ofsecurity, and well-developed communication skills. A carefully crafted campaign to educatepublic opinion is essential. Each reform package must be screened from the standpoint notonly of economics but also of political feasibility. Political considerations frequently mayrequire alterations in the technical design of proposed economic policies aimed at elicitingthe cooperation of all parties (within government and outside) who have the power to blockthe reforms. In some instances such cooperation can be obtained only by partiallycompensating the short-run losses of well-organized interest groups. It may also be necessaryto phase in the reforms over time to avoid the coalescing of different groups opposed tovarious elements of the reform package. But once a decision is made to tackle a particularissue, there is much to be said for moving swiftly. Whatever the political tactics, they mustbe chosen carefully by the top leaders and their trusted aides.

A special aspect of the policy process is the interaction between African governments andmultilateral agencies (the IMF and the World Bank, in particular) as well as bilateraldonor agencies. These foreign actors have joined the policy circle in many African countries.During the 1980s, the World Bank and the IMF set the agenda and the main parameters ofthe policy negotiations. Indigenous policy capacity remained very weak. Many governmentspassively accepted the initiatives pressed by international agencies. Desperate to obtainloans conditional on policy agreement, governments in many instances signed the documentswithout securing firm domestic commitments. Not surprisingly, the implementation of theseagreements frequently proved to be a stumbling block. Leaders and senior civil servants didnot feel a sense of "ownership" about the reforms. In many cases these reforms had not beensufficiently adapted to the local institutional, social, and political scene.

It is important that African governments reacquire primary responsibility for defining andimplementing their economic policies. They must exercise initiative in identifying policyissues, in preparing alternative policy programs, in selecting policy options that areeconomically and politically feasible, in managing the round of negotiation with foreignpolicy actors, and finally in monitoring as well as correcting policy measures as they arebeing implemented. The locus for decision and action has to shift back to the capital city ofeach African government. There is no other way to make policy decisions that are credible,workable, and sustainable. The IMF and the World Bank have an important policy role toplay in Africa, but it is a significantly different role than the one they played during thepast decade, as I will explain later in the chapter.

I conclude this section by putting forward the view that individual countries have alimited capacity and willingness to undertake reforms at a point of time. This capacity andwillingness are determined by the quality of the top leadership (and of the civil service) aswell as by the social stratification and organizational effectiveness of pressure groups.Furthermore, leaders change in response to adversity or because they have learned fromhistory or because they are exposed to new ideas. In turn, leaders can mobilize the civil

6. The ascendancy of President Chung-Hee Park in South Korea in the early 1960s and hisdetermination to undertake serious economic reforms led to improved organization of high-qualityprofessional work in the policy area, including the growth of the Korean Development Institute in therole of advisor to the Economic Planning Board. Similarly, President Raden Suharto in Indonesiaelicited a major professional contribution into economic policy in the late 1960s from the group ofIndonesian economists educated in Berkeley, often referred to as the "Berkeley Mafia."


service though training, reordering of incentives, and reorganizing of the governmentmachinery. In these ways the capacity and the willingness of a country to undertake reformschange over time. A good match is important between such capacity and willingness on theone hand and the kind of reform that is attempted. A country with limited capacity orwillingness is not likely to succeed in adopting reforms that are demanding in terms ofpolitical management and administrative competence. These requirements of reforms dependnot only on how ambitious the goals are, but also on how good the technical design is of theproposed measures.

Improving the Technical Design of Reform Packages

As examples in Chapter 5 showed, weaknesses in the technical design of reforms canexacerbate the problems of implementation and reduce the positive net impact on theeconomy. It follows that improvements in theory, applied analysis, and statistics can yielda more accurate diagnosis of economic problems as well as a better recipe for what should bedone. This section focuses on the technical aspects of policy design within the relativelynarrow analytical framework of professional economists. Once technical contributions havebeen made, policy proposals need scrutiny from the much broader perspective of politicaleconomy. In the interest of facilitating the political management of reforms, it may benecessary to alter the scope, timing, and sequence of policy packages, derived on the basis oftechnical analysis.

A brief word about developments in economic theory may be useful here. In Chapter 3, Isketched very briefly the mnain ideas that legitimized state interventions to correct "marketfailures." The tide turned in the late 1960s and 1970s. By the beginning of the 1980s manyeconomists had lost confidence in the state. Milton Friedman (1970), the founder of themonetarist school and the champion of a minimal state, had displaced in the economicslimelight John Maynard Keynes (1936), the chief advocate of active fiscal and monetarypolicies. The views on economic policy of Harry Johnson (1971), Peter Bauer (1972), andJagdish Bhagwati and T.N. Srinivasan (1975) were a formidable attack on the doctrine ofimport substitution via forceful protection associated with Raul Prebisch (1959). Publicchoice theorists, such as James Buchanan and Gordon Tullock (1962), rejected the view of themonolithic state and began to analyze how the separate interests of different state agentsaffected economic policy. Ann Krueger (1974) enunciated the theory of "rent-seeking,"according to which economic agents compete legally and via the black market to acquirewindfall gains created by government controls. There was a virtual "counterrevolution' indevelopment philosophy (Toye 1987).

Growing skepticism about the state was not confined to the political right; it was sharedin some measure by Marxian and other radical economists (Killick 1989, p. 16). "There was atendency for the reaction against dirigisme to go too far.... There has thus occurred a partialrehabilitation of the state-but one which has left much scope for debate about thedesirable extent of its involvement in the economy" (Killick 1989, p. 32). The most recentemphasis of the economics profession is on applying cost-benefit techniques to questions aboutwhat the state should do and how it should do it. It is not easy to apply these techniques,of course, but the general approach is useful. "Market failures" should be balanced against"bureaucratic failures."

What should be included in a policy package? This is a critical question. Governmentshave multiple economic goals, and according to the well-known theorem expressed byTinbergen (1967), the number of policy instruments must equal the number of policy targets.Different parts of the economy are linked to each other as are different policies. Only ageneral equilibrium approach to policy design can deal adequately with multiple goals andmanifold linkages. This approach tends to favor comprehensive policy packages thatsimultaneously address interrelated problems and objectives. Such packages can, in principle,produce mutually reinforcing, synergistic effects and thereby accelerate and amplify the


positive impact of reforms.Experience in Sub-Saharan Africa and elsewhere suggests that comprehensive packages

are very difficult to design and even more difficult to implement. The political andadministrative problems have already been discussed. The technical problem is uncertaintyregarding what is wrong with the economy, what the precise linkages between differentparts of the economy are, and what the economy's response to different combinations ofpolicy instruments will be. Furthermore, multiple goals, some of which conflict with eachother, require the construction of a set of weights with which to evaluate competing policyproposals. This is beyond the technical realm. The derivation of these weights involvesvalue judgments, which once again brings us back to political matters.

All these considerations make essential a selective approach to policy change. Thecriteria for selection that can help define the scope of policy packages and their sequence isfar from clear. Theory is weak in defining the transition between the existing policy regimeand the desired one. If there is a unique optimal path that connects the two regimes, nobodyknows how to chalk it out. It is perfectly possible, of course, that reasonable progress can bemade along a number of different paths and that the search for the optima will not be veryrewarding. According to Streeten (1987, p. 1473), "We completely lack a handbook for reform-minded Prime Ministers and Presidents who would like to know how to manage thetransition to a better society." In the absence of theoretical guidance, policymakers areobliged to "learn on their feet." The following subsections will review the operational issuespertinent to a number of aspects of stabilization and structural adjustment.

Stabilization Programs

A remarkable feature of policy experience during the 1980s was the high incidence ofbreakdowns in stabilization programs. Undoubtedly, a part of the explanation for thisphenomena is to be found in the political realm (Zulu and Nsouli 1985). As noted earlier,some governments wavered in their political comrnitment to stabilization targets or signedthe documents simply to get desperately needed money from the IMF or debt relief from theParis Club. That was not the whole story, however. Helleiner (1986c) discusses problemsraised by the short time horizon within which stabilization was attempted by reducingbudget and current account deficits. Taylor (1988) points out that IMF financial programmingwas based on extreme monetarist and neoclassical assumptions. I mentioned these problems inthe context of Zambia and Malawi in Chapter 5. In most SSA countries ambitiousstabilization targets required substantial cutbacks in government expenditures and in importssince rapid expansion in public revenue and exports was not feasible. The cutbacksapparently were determined on the basis of macrofinancial models and did not takesufficient account of implications for the real economy. The predominantly top-downapproach created numerous tensions at sectoral and micro levels during implementation. Insome instances, governments managed to adhere to the stabilization targets but only at theexpense of debilitating shortages of key intermediate goods and services. These shortagesprevented the full use of plant and equipment and caused the productivity of governmentextension workers, teachers, and other personnel to decline. In other cases, tensions generatedby these targets were so severe that implementation was derailed.

It appears that the design of stabilization programs agreed to by individual SSAgovernments and the IMF did not reflect the specific characteristics of these countries.Zambia, for example, was much more seriously affected by exogenous shocks and policydistortions than was Malawi or Mauritius. It was clear that stabilization was a differenttype of problem in the three cases. As discussed in Chapter 5, it was relatively easy torestore financial balance in Mauritius by some demand restraint combined with a realdepreciation in the exchange rate aimed at restoring the momentum of growth in themanufactured export sector. The EEC sugar quota cushioned the impact of fluctuations in theworld market. Malawi's situation differed. There was no ready-made apparatus forobtaining a substantial supply response in quick order. Furthermore, as a landlocked country,


Malawi was particularly vulnerable to exogenous shocks. In Zambia the inherent problemswere deep rooted-the extreme specialization in exports of copper, the institutional andphysical difficulties in producing copper, and the protracted neglect of the traditionalagricultural sector.

Despite these differences in the three countries, stabilization programs in Zambia,Malawi, and Mauritius were based on the same approach, featuring a short time horizon anda heavy emphasis on demand restraint. As noted in Chapter 5, the chosen approach workedfairly well in Mauritius, had mixed results in Malawi, and was abandoned in Zambia.

Future attempts to stabilize should be tailored to the individual circumstances of eachcountry. Taylor (1988) suggests the need to improve the Polak model (which has underpinnedIMF stabilization since 1957) by synthesizing monetarist and structuralist dimensions. Eachcountry assessment should answer the question "Is inflation likely to be reduced by austerity,and with what effect on output by sector and over what time horizon?" (p. 160). Targets forstabilization should be determined on the basis of a judicious balance between curtailment ofdemand and expansion of supply. Of course, demand can be reduced sharply through fiscal,monetary, and other policy instruments. This may be appropriate in some circumstances,particularly when the problem is very largely one of an overheated economy following asubstantial real economic expansion for a considerable time. It may be much more difficult,however, to sustain austerity, particularly if the economy has been in a downward spiral.Even if austerity is sustained, it is likely to affect adversely the utilization of existingcapacity and the pursuit of long-run development goals. It may be desirable, therefore, toattempt stabilization over a longer period and to rely more on increases in production andexports and less on a cutback in investment or consumption. Judgments will have to be maderegarding the time required to solve specific investment, production, or marketing problems.Lags separating the initiation of policy measures and their full impact can be of varyingduration depending on the nature of underlying problems and the stage of institutional andeconomic development. It is important that stabilization programs are built on reasonableassumptions about economic growth and that foreign aid requirements are calculatedaccordingly. Implicit in the recommended approach is the premise that donors active in Sub-Saharan Africa will respond to an authoritative assessment by the World Bank and the IMFof aid required to support reform efforts in individual countries. I will return to this issuelater in the chapter.

A second problem after the choice of time horizon concerns the treatment of shocks.Typically, stabilization programs are based on specific assumptions regarding a number ofvariables including commodity prices and the amount of foreign aid. Experience shows thatevents have falsified these assumptions very frequently. Weather conditions, internationalprices, and donor behavior have been difficult to predict. These disturbances havecomplicated macroeconomic management because many countries have not had recourse toadequate reserves of foreign exchange or sufficient access to short-term credit with which tooffset shortfalls in predicted farm output, international prices, and aid. Against thisbackground, it is reasonable to question the usefulness of very precise, time-bound targets thatare intended to instill monetary and fiscal discipline. Such targets can introduce unnecessaryrigidity in macroeconomic management. Efforts should be made to establish adequatecontingency arrangements enabling governments to cope with the considerable uncertaintythat bedevils their financial operations. These arrangements can take the form of anexpanded and greatly liberalized compensatory finance scheme or the adoption of targetzones instead of single-point targets. Action was taken by the IMF in 1988 to provide funds tocover part of the adverse impact of unfavorable deviations in export earnings, import prices,and international interest rates. These deviations are measured in relation to baselineprojections. The IMF has taken a step in the right direction, but whether it is sufficientremains to be seen.

The short horizon syndrome characteristic of recent reform efforts in Sub-Saharan Africaadversely affected not only stabilization policies but also agreements with the World Bankon structural adjustment measures. The Bank expected that policy and institutional moves


over perhaps four to five years would be sufficient to restore normalcy. In retrospect, thisexpectation was far off the mark in most of Sub-Saharan Africa. Right or wrong, the premisethat economies could be put back fairly rapidly on the long-run growth path generated apreference for prescriptions that could be administered quickly. Priority was attached torestoTing financial balances, and it was assumed that stabilization would be followed by asatisfactory rate and improved patterns of economic growth. Unfortunately, there was awidespread tendency to eschew systematic analysis, to treat superficial symptoms instead ofunderlying causes, and to adopt, in some instances, the currently fashionable recipe. Thesetendencies in the four major areas of reform first discussed in Chapter 2 (the public sector,exchange rate policy, agricultural policy, and manufacturing and trade policy) will beexamined in the following paragraphs.

Public Sector Reform

Reduction of fiscal deficits can be achieved by quick, across-the-board expenditure cutsand ad hoc revenue hikes. This short-term approach can yield fast results, but it willadversely affect long-run growth and socioeconomic development. These negative side effectsmay call into question the sustainability of the reforms. Avoiding these side effects,however, requires deeper studies, much more professional expertise, and a longer timeperspective.

A good example is the restructuring of a public expenditure program. It is relatively easyto cut outlays by a fixed percentage across the board. It is much more difficult to determinethe relative priority (from the standpoint of economic growth) of capital and recurrentoutlays in different sectors. Equally or more difficult is designing cutbacks in publicexpenditure and at the same time protecting or even improving the provision of services topoverty groups. Efforts were made during the 1980s to determine expenditure priorities bythe fundamental rather than ad hoc approach, but many formidable obstacles wereencountered. There were very serious data difficulties in many instances, but scarcely anygovernments set in motion a statistical improvement program to deal with this problem. Itwas only in 1988 that the Social Dimensions of Adjustment Project was launched to improvethe information base for designing structural adjustment programs that would minimize theadverse impact on the poor. (The project is financed by UNDP, the World Bank, and others).Furthermore, the effort to establish expenditure priorities and to adhere to them ran intocomplex institutional and procedural bottlenecks in many countries. These issues can betackled only by a long-term program supported by the country's top political leadership.

Prioritization of public expenditure programs raises important methodological questions.Existing techniques of cost-benefit analysis are not capable of establishing the relativeattractiveness (from the growth standpoint) of projects and programs in different sectors(Baum and Tolbert 1985, p. 443). The use of the discount rate in such analysis tends todiscourage investments with long-term benefits and to encourage those with long-term costs.During an economic and financial crisis, the discount rate increases to signal the accentuationof resource scarcity, thereby enhancing the preference for short gestation and quick-yieldinginvestments. Measures to secure long-term goals, such as building human capital andprotecting the environment, may have to be postponed, according to this logic. Suchdisplacement of important matters by urgent ones is tolerable in the context of a brief crisis,but it becomes a matter of rising concern as the abnormal period of austerity gets moreprotracted. Although cost-benefit techniques serve an extremely useful purpose, they cannotbe the only basis for allocating resources between activities to build human capital and otheractivities. This decision is a strategic one; once it is made, cost-benefit analysis can beapplied to refine intrasector and project-by-project allocations.

Sub-Saharan Africa has already experienced a long, difficult period in which themomentum of growth has been rudely disturbed and in which vulnerable groups have


suffered greatly.7 The entire process of capital accumulation (physical and human) has beenundermined with grave implications for long-run development. It would be unfortunate toallow this setback to persist even longer. This point will be discussed later in this chapter inthe section on the international framework.

Many SSA governments have resorted to a wage freeze for government workers. Undercertain circumstances this measure can be justified for a brief spell, but in Sub-Saharan Africait has gone far beyond that. In many situations it is no more than a palliative. While thefreeze has helped contain government expenditures, it has led to moonlighting by governmentworkers. Their incentive to work hard has been undermined. Moreover, the government'scapacity to fill vacancies at high levels in the civil service has been reduced. A permanentsolution will require a thoroughgoing assessment of how much of the government work forceis redundant and what is the optimal pattern of wage differentials for workers with varyingskills. It will also be necessary to evolve a policy that combines compensation with training,improved management practices, and career development aimed at securing an efficient civilservice.

Parastatal reform remains a difficult area. Many SSA governments in the 1970sestablished state enterprises to correct market failures and improve income distribution.However, experience did not confirm the political and administrative capacity of manyAfrican governments to intervene effectively. It is all too easy now to adopt the oppositeparadigm and argue that government should play a minimal role, but such a shift cannot bedefended on professional grounds. The technical problem involves predicting the futurebehavior of governments on the one hand and of private agents on the other. There is norobust method for making these predictions, and it is clear that there exists much diversityas far as efficiency is concerned both in the public and private sectors of individual countries.It follows that pragmatic solutions have to be tailored to the individual case.

In practice, governments in the sample countries showed a marked preference forrehabilitating state enterprises instead of divesting them. Uganda sold seven parastatals,Tanzania sold three, and five other countries sold one each. Operations of state enterpriseswere suspended in some cases, but these firms retained a legal status-a process labeled"withering away" by Berg and Shirley (1987, p. vii). Governments were reluctant to sellprofitable enterprises, and it was difficult to negotiate terms of sale for loss-making firms.Government also tended to exclude buyers from certain ethnic groups or from overseas. Thevolume of activity under the heading of rehabilitation was much larger and consisted ofdiagnostic studies, preparation of action plans, financial restructuring, upgrading ofmanagement, and establishment or strengthening of government agencies charged withsupervision of parastatals.

According to Nellis (1986, p. xi), these reforms retain a "fluid and experimentalcharacter." He also warns that too much should not be expected from the so-called "contractplans" (CPs) negotiated between governments and individual parastatals (Nellis 1988, p. ix).Contract plans are not good instruments for restoring normalcy to crisis-prone parastatals.Even with respect to parastatals in good health it is difficult to find "the proper balancebetween empowering the managers through the CPs, and at the same time endowing anoversight agency with sufficient 'clout' to enforce the CP" (Nellis 1988, p. ix). It is obvious

7. Sub-Saharan Africa has 280 million people who are absolutely poor. Life expectancy fell in ninecountries between 1979 and 1983, and primary education enrollment ratios declined in twelve.According to the World Bank (1989, p. 4), "For many developing countries there was simply noalternative to adjustment, and some groups did benefit (including some relatively poor groups such assmall farmers in Africa). But many vulnerable groups inevitably suffered as austerity reduced theavailability and affordability of social services, eroded real wages, and reduced public sectoremployment. The addition of new groups to the absolute poor and the persistence of absolute povertyin countries suffering economic crisis increased the urgency of the problem. By the mid-1980s broadexperience with adjustment made it clear that, given the time and effort required to turn deeplytroubled economies around, it would be morally, politically, and economically unacceptable to wait forresumed growth alone to reduce poverty."


that development professionals do not possess clear-cut solutions to problems plaguing theparastatal sector.

A forceful structural adjustment program for ministries and parastatals in many countriesis likely to lead to sizable retrenchment. The newly unemployed (or the "new poor" as theliterature describes them) are unlikely to soon find gainful employment producing exports orimport substitutes unless the overall economy recovers and expands rapidly. Instead,transitional unemployment and underemployment probably will increase. Although someinnovative credit, training and public works programs have recently been launched, testedrecipes that government can use to facilitate the reabsorption of these workers are not yetavailable (Zuckerman 1989).

Exchange Rate Reform

Although in 1980 many policymakers in Sub-Saharan Africa distrusted exchange rates asan instrument for stabilization and structural adjustment, the past eight years have seen agreat deal of experimentation in this area. Several countries succeeded in reducingsignificantly the extent of distortion in this key price, while others encountered asubstantial measure of frustration. Different postures emerged in different countries.Ethiopia, for example, did not alter either the currency peg or the parity value of the birr.Instead, it tried to maintain international competitiveness by heavy reliance on demandmanagement through fiscal and monetary policies. A group of countries changed both currencypegs and parity values. In Kenya, Zimbabwe, and Malawi the aim was to adjust thecontrolled price of foreign exchange to compensate for higher domestic inflation comparedwith that in their respective trading partners. A moderate measure of depreciation in thereal exchange rate was also achieved in these countries. Madagascar and Tanzania adopted,after some hesitation, a more forceful policy of depreciating the real exchange ratesubstantially over a period of time in order to correct large distortions that had accumulatedduring the preceding period.8 Uganda, Zambia, Somalia, and Zaire conducted even moreradical experiments involving flotations, auctions, multiple exchange rates, and other exoticfeatures. Several of these arrangements were aborted.9

The key issue in designing reform of exchange rates is not whether nominal devaluationby itself is a solution but rather what role flexible exchange rates can play in a packagethat also includes other policies. Experience in Sub-Saharan Africa suggests that nominaldevaluations do create some extra inflation. Over the medium term, the magnitude of realdevaluation is usually substantially less than the amount of nomninal devaluation. Duringthe 1980-86 period, sample countries varied greatly with respect to how much nominaldepreciation they needed to secure 1 percent depreciation in the real effective exchange rate.The ratio was 10 or less in Mauritius, Kenya, Malawi, and Zimbabwe. By contrast, thecorresponding ratio was more than 10 but less than 40 in Madagascar, Zambia, Somalia, andZaire. The ratio exceeded 100 in Uganda and the Sudan. The extreme case was Tanzania,where very large nominal devaluations during the early 1980s were accompanied by asubstantial appreciation in the real rate.The policy lesson is that a one-time nominal devaluation does not generally produce thedesired real change unless a very restrictive fiscal, monetary, and wage policy is also in

8. The real exchange rate depreciated by 24 percent from 1983 to 1986 in Madagascar and by 31percent in 1986 in Tanzania. Further progress was made in subsequent years in both cases.9. The real exchange rate depreciated by 87 percent in Uganda during the 1980-84 period, but thismovement was reversed subsequently. The corresponding depreciation in Zambia was 65 percentbetween 1982 and 1986, followed by some appreciation in 1987 and 1988. In Somalia the realdepreciation was 69 percent between 1984 and 1987, followed by a reversal. In Zaire the realdepreciation was 69 percent during the 1983-87 period, followed by a very slight movement in theopposite direction in 1988.


effect. Such an austere stance can undermine the momentum of production and investment.Macroeconomic policy managers have to minimize these adverse side effects by sensitivehandling of policy levers amidst a large measure of uncertainty. What may be required inthis context is a flexible exchange rate regime and a determination to achieve a specifiedreal depreciation over a well-defined period.

Devaluation can exacerbate budgetary problems. The net impact on the budget in theshort run depends on the structure of taxes and expenditures. Pinto (1988), for example, hasargued that the differential between the predevaluation official exchange rate and theblack market rate is an implicit tax on exports. Whenever government is a net buyer offoreign exchange, a reduction of the differential (as a result of devaluation) would imply aloss of government revenue, and in the absence of fiscal reform it would mean moreinflationary financing of the budget deficit. However, this outcome is not inevitable.Simultaneous action can be taken at the time of devaluation to safeguard the government'sfinancial position by raising taxes or reducing expenditures.

Assuming that a country succeeds in depreciating its real exchange rate, the next questionis the extent of the economy's response to the new structure of prices. Under Africanconditions economic agents are constrained by a number of nonprice variables such asphysical, technological, and institutional obstacles. These nonprice factors weaken short-run,supply-side responses to real devaluation. The length of lags and the strength of the long-runresponse are a function of what other measures are included in the policy package and howeffectively they are implemented. It is very important in Sub-Saharan Africa that efforts to"get prices right" through devaluation and other policies are accompanied by projects tofacilitate investment, technological improvements, managerial changes, and institutionaldevelopment.

If real depreciation and supplementary action to tackle nonprice constraints lead to alarge increase in the volume of exports of traditional commodities from producing countries,then world prices may fall to the detriment of all of Sub-Saharan Africa. Internationaldemand for these commodities is price inelastic. This "adding-up problem" must beconsidered in designing policy packages. It may be necessary to raise taxes on such exports(simultaneously with devaluation) to prevent undesirable increases in supply. This was donein Mauritius with respect to sugar. Alternatively, agreements with suppliers from outsideSub-Saharan Africa can be negotiated to restore the historical share of African producerswithout causing a slump in world prices. History suggests that such agreements are difficultto arrange. In any event, a dynamic exchange rate policy remains an essential prerequisitefor exploiting present and future comparative advantage (Gulhati, Bose, and Atukorala 1986,p. 413).

It is not easy to calculate how much real depreciation is required. The rate should reflectthe scarcity value of foreign exchange. This will be the greater (a) the more restrictive isthe import trade regime and thereby the larger is the suppression of the demand for foreignexchange; (b) the higher is the projected debt service and the higher the level of arrears; (c)the less favorable is the prospect for the international terms of trade; (d) the smaller is theprospect of sustaining or raising levels of official development assistance on a long-termbasis; (e) the higher is the minimum rate of growth of GDP acceptable to policymakers andtherefore the required increase in imports. Given the difficulty of translating theseconsiderations into a number, policy has to be on a trial-and-error basis. A step-by-stepapproach will be necessary involving monitoring and evaluation of an initial rate adjustmentfollowed by the design of the next move. There is not much agreement on the speed ofadjustment. Those who favor gradual adjustment emphasize the structural rigidities ofAfrican economies, including the political economy costs of making drastic changes in sosensitive a price as the exchange rate. By contrast, those who favor a "quick fix" stress theneed to alter expectations of economic agents and to establish the credibility and durabilityof the new policy.

Those who believe in market-determined exchange rates do not have to answer questionsregarding how much and how rapidly to devalue. In fact, not having to find answers to these


questions is part of the attractiveness of deregulation. The government can eschewresponsibility for the rate by establishing the market for foreign exchange and therebydepoliticize the issue. Attention then shifts to the following problems:

* How can a government move from the fixed, controlled exchange rate to a market rate?What should be the transitional steps, and how long should the transitional phaselast? Who should have access to the foreign exchange market and for what kind oftransactions? How much foreign exchange should be sold and how much put in reserves?

* What measures can be taken to facilitate domestic currency borrowing by importantgroups (such as small farmers) so that they can buy enough foreign exchange? How canthe creditworthiness of such borrowers be improved? What can be done to raise theefficiency of financial intermediaries?

* What can be done to help small farmers, and other groups, cope with fluctuations inexchange rates? African countries are very vulnerable to shocks. The market exchangerate transmits all these fluctuations to economic agents regardless of how temporarythey are. These fluctuations will affect adversely savings, investment, andtechnological changes unless economic agents can protect themselves.

The appropriateness of market-determined rates as a permanent feature of SSA economicpolicies is questionable, however. Even in mature industrialized countries freely floatingrates have been greatly criticized; in Sub-Saharan Africa their suitability is much moresuspect. It will not be easy to establish forward markets or alternative insurance schemesthat can effectively insulate small, unsophisticated economic agents from the risks offluctuations. Official transactions that are not responsive to market phenomenon dominatethe capital account of SSA countries. For example, trends in international and domesticinterest rates are not likely to affect the net inflow of ODA or official loans on conventionalterms. The danger of capital flight on private account from SSA countries remains large, andthis will impede liberalization of the capital account. The exclusion of the capital accountfrom the foreign exchange market will destabilize exchange rates further. Finally, thethinness of foreign exchange and financial markets in these countries will introducedistortions and possibly the exercise of monopoly or monopsony power in the determination ofexchange rates. Therefore, while market-determined exchange rates can be an effectivetransitional arrangement, they should not be thought of as durable.

Agricultural Policy Reform

Reforms during the 1980s touched on many aspects of agricultural policies and institutions,but their main focus was on improved pricing. In addition, foreign donor agencies pressuredSSA governments to liberalize agricultural markets by redefining the function of parastatalsand by extending the operational scope of private agents. Some progress was made on bothfronts, perhaps more on getting the prices right and less on liberalization. Issues of designplagued both types of reform activity, however.

Policy attention focused on official producer prices. However, little is known about therelationship between these official prices and the prices farmers actually receive. Fordomestically consumed food crops, the relationship may well be a distant one, given theprevalence of parallel markets and the difficulties of enforcing trade through officialchannels.10

10. Green (1987) states, "A rough estimate for SSA as a whole (country positions would vary widely) isthat 75 percent of domestic food is self provisioning ('subsistence') or very local non-official channeltraded. Of the remaining 24 percent about 13 to 15 percent is of crops/animals of which (generally or insome countries) there were no official prices, another 5 percent is parallel marketed and at most 7percent is directly affected by official prices. Trying to raise output by acting on 7 percent of output


Many observers have concluded that while price policies are important, they are not thekey variables for explaining variations in overall agricultural production. Farmers, of course,respond to relative prices of competing crops, but their aggregate response to movements inthe terms of trade between farm and nonfarm products is quite small particularly in theshort run. Even in the long run, the aggregate supply elasticity of agriculture with respect toprices is estimated to be within the range of 0.2 to 0.5 in countries at the SSA stage ofdevelopment (Chhibber 1988, p. 45). By contrast, the corresponding elasticity with respect tononprice variables (feeder roads, other transport, research, extension, water supply, credit,social services) is around 1.11

Cleaver (1985, p. 29) concludes that reform packages should focus first on relaxingnonprice constraints and second on price improvements. My own conclusion is somewhatdifferent. Governments should pursue simultaneously reforms affecting both prices andnonprice variables. On the one hand, excessive emphasis on prices is not likely to be veryrewarding; in fact, it could be counterproductive since sharp changes in price relationsthreaten the existing income distribution. On the other hand, concentrating on projects toimprove nonprice variables is likely to be wasteful in the presence of serious misalignmentsin the structure of incentives. Indeed, this is one of the lessons of experience during the 1970s.

Somalia, Zaire, Madagascar, and a number of other countries made progress during the1980s in liberalizing their agricultural markets. However, the pace of change was slowerthan expected. Apart from ideological resistance and pressure group activity to blockreforms, Berg (1987, p. 24) points out that the pay-off to liberalization over the short andmedium term may not be very large, if parallel markets already provide reasonablealternatives to the official channel. Furthermore, improvements in the efficiency of privatetraders in a decontrolled system cannot be taken for granted. Private traders will tend tooperate in relatively lucrative market segments, and governments will still be obliged toprovide services to scattered farmers who produce small marketable surpluses.

Another important aspect of agricultural policy design concerns measures required toprotect (or preferably to enhance) the primary incomes of poor farmers, especially womenfarmers. Such measures are intended to increase their access to land, institutional credit,transport, extension services, and markets. These concerns enjoyed a great deal of prominencein the 1970s under the rubric of integrated rural development. They have now beenrearticulated in professional seminars and international conferences under the heading ofstructural adjustment.

Lying behind these issues is the general problem of defining the role that governmentshould play in the agricultural sector in individual SSA countries. This is fairly easy if oneassumes that government can intervene effectively. Lele and Christiansen (1988, p. 20) arguethat governments should correct market failures by guaranteeing producer prices, ensuringfood security, providing adequate infrastructure, reducing risks facing small-scale farmers,and promoting research and extension. They also argue in favor of governments promotingequity. Experience during the 1970s and 1980s did not confirm, however, the capacity of SSAgovernments to intervene effectively, thereby bringing into question their future role inagricultural development. What the new strategy should be for the period ahead is not easyto define on the basis of technical considerations only.

(higher in surplus or good years and lower in drought or bad years) is hardly likely to be spectacularlyeffective."11. The coefficients cited by Chhibber (1988) are drawn from a number of sources. Studies made inIndia covering 1954-55 to 1977-78 concluded that aggregate long-run price elasticity of supply exceeded2. When nonprice variables (percentage of area irrigated was used as a proxy) were introduced into theregression equation, the price elasticity coefficient fell to 0.30 and the coefficient for irrigation was equalto 1. Another study by Peterson covering 53 countries yielded a price elasticity of 1.7. This coefficientdeclined to 1.27 when a research variable was included and further to 0.97 when an irrigation variablewas included.


Manufacturing and Trade Policy Reforms

This section will focus on attempts to liberalize trade policies by Mauritius, Kenya, andZambia. Such efforts, combined with macroeconomic policy changes, were a prominentfeature of reforms affecting the manufacturing sector in these countries. The reforms were andstill are controversial. Some critics say that dismantling quantitative restrictions will causea flood of imports, put pressure on foreign exchange reserves, or lead to excessive externalborrowing. Another argument often offered against dismantling them is that it will divertscarce foreign exchange to imports of nonessential consumer goods. Skeptics also argue thatliberalization of imports that compete with locally produced manufactures will injureAfrican firms and lead to deindustrialization. Furthermore, critics say that export promotionis outmoded. It was less promising during the late 1970s and 1980s in the face of slow growthin world trade and rising protectionism than it was in the 1960s and early 1970s when globaltrade was rising very rapidly. African manufacturers, in particular, will find it verydifficult to penetrate world markets, it is argued.

The three cases of trade policy reform in Mauritius, Kenya and Zambia illustrate a widerange of designs and outcomes. In Mauritius trade policy changes were not initiated untilafter the economy had been stabilized. Chapter 5 showed that the initial results of theseincomplete reforms were promising. Kenya abolished import prohibitions before the economyhad been stabilized but government could not implement the agreed program for liberalizingitems subject to QRs, as explained in Chapter 4. In Zambia import prohibitions wereabolished and restrictive import licensing dismantled all at once in the middle of an acutebalance of payments crisis. Chapter 5 showed that these changes did not survive theabandonment of the entire reform program soon after imports were liberalized.

History is full of examples of aborted trade policy reforms. Some setbacks, such as thoseexperienced in Kenya and Zambia during the early 1980s, are likely in other Africancountries, but such episodes should not be allowed to engender wholesale pessimism. Instead,policymakers should learn from these experiences and design improved reform packages. Badmacroeconomic management, such as lax fiscal and monetary policies, can doom trade policyreform. Appropriate behavior of the real exchange rate is crucial to successful reform.Realistic exchange rate valuation is a necessary condition for expanding exports andrestraining imports, particularly when quantitative restrictions are being removed andimport tariffs are being rationalized.

Exogenous and policy factors that produced rapid export expansion in Mauritius whileliberalization was taking place helped in the implementation of trade reforms. In Kenyathe 1982 foreign exchange crisis undermined trade policy reforms. In Zambia continuing acuteshortages of foreign exchange led to the abandonment of the reform. The skeptics are right topoint out the danger that trade liberalization may cause a flood of imports. However, suchan outcome is not inevitable. What is required is an appropriate macropolicy frameworkwithin which to launch trade liberalization. Furthermore, reducing the protection level doesnot necessarily mean tolerating an undesirable diversion of scarce foreign exchange to importsof nonessential goods. A combination of import tariffs and equivalent taxes on locallyproduced nonessential consumer goods can avoid such a diversion of foreign exchange.

There is considerable merit in pursuing a preannounced path of trade policy reform inAfrica. It is important to classify existing firms into three groups: those that cannot be madeviable even after restructuring; those whose existing capacity can be restructured profitablyeven though replacing or expanding capacity would not be a good idea; and those that arealready competitive or that can be made so through profitable restructuring (including theupgrading of skills and management practices) and new investment. Reform involving asudden reduction in protection can lead to many casualties, including disruption of enterprisesin the latter two categories. A gradual approach would seem to be the wiser course. Providedpolicy signals can be made clear and credible, firms should have sufficient time to adjust.Technical assistance and capital should be made available on reasonable terms to assistfirms in the latter two categories to restructure their operations. Of course, firms in the first


category will have to be closed down, but the resulting sociopolitical problems should bedealt with through compensation, relocation, and retraining. However, gradualism shouldnot permit doubt about the commitment to reform. Transitional policies should not degenerateinto attempts to fine-tune every move. In Africa particularly, such fine-tuning is notpractical, given the acute shortage of data and professional skills.

Some expansion of manufactured exports can be expected as a result of trade policy andexchange rate reforms. However, rapid increases in exports of manufactured goods areunlikely in the near future. Unlike the newly industrializing countries, SSA countries are notat the stage of economic development where they already have sufficient physical andhuman capital as well as markets and institutions that function reasonably well. Measureshave to be taken to improve these essential components of the manufacturing sector. Reformsin trade policy, however, do not have to wait until these developments have alreadyoccurred. By increasing competition, import liberalization can help improve the overallefficiency of investment in and production of import substitutes, thereby preparing the groundfor a later round of expansion in manufactured exports. It is essential, however, forgovernments to dismantle domestic barriers to competition at the same time that they reformpolicy affecting foreign trade.

Improving the International Framework

Economic reforms in SSA countries during the 1980s were stymied by a hostileinternational environment (see Chapters 4 and 5). Adverse trends in the international termsof trade, fluctuations in world interest rates and exchange rates, and a setback in foreigncapital flows accentuated the difficulties for countries trying to carry out reforms.Organizations like the Paris Club, the IMF, and the World Bank were not fully equipped toperform their respective roles effectively. I do not intend to develop a full-fledged analysisof the current international framework or its future reform. This is a topic for another book.A complete treatment of the economics and politics of this subject would be very desirable.My purpose in this section is to describe briefly the characteristics of the internationalframework in a way that highlights issues that deserve further consideration.

Economic Policies of Major Powers

A major change occurred at the end of the 1970s in the orientation of economic policies ofthe Group of Five (G-5)-the United States, the United Kingdom, France, Germany, andJapan. In response to relatively high inflation, these major powers in late 1979 adopted anaustere monetary policy that, together with the impact of the second oil price shock,precipitated a severe and protracted recession. The rate of unemployment in Europethroughout the 1980s remained at a very high level by post Second World War standards.Real interest rates rose sharply, particularly in the United States, which became theworld's largest debtor. The real exchange rate of the U.S. dollar appreciated veryconsiderably in the first half of the 1980s and then started to depreciate until the middle of1989. Only very limited progress was made to synchronize G-5 policies with the aim ofmanaging the world economy in the interest of all participants in the international system.Most major players placed much more emphasis on domestic concerns than on global ones.

Sharp fluctuations in interest and exchange rates of the major industrialized countriesgenerated a great deal of uncertainty for SSA governments and complicated their problems ofmacroeconomic management. Real prices of primary commodities exported by Sub-SaharanAfrica fell very sharply during the 1980s, reflecting a variety of factors, includingrelatively low rates of real growth in GDP in the industrialized countries. Compared with a1979 projection of 4.2 percent per year real growth in GDP of industrialized countries from1980 to 1990, the most recent assessment (which takes account of actual growth in the 1980-87period) is only 2.5 percent (World Bank 1988a, p. 14). Another relevant factor was the


Common Agricultural Policy of European Common Market countries that discriminatesagainst imports of primary commodities, including those originating in Sub-Saharan Africa.Similar trade obstacles exist in the United States and Japan.

Foreign Aid and Debt

The volume of foreign aid to Sub-Saharan Africa, which increased very briskly duringthe 1970s, tended to stagnate during the early 1980s (Gulhati and Nallari 1988). Donorsbelonging to the Organization for Economic Cooperation and Development (OECD) tried toredirect official development assistance to governments in Sub-Saharan Africa undertakingintensive reforms, but these efforts were insufficient. A variety of nondevelopmentalconsiderations (geopolitical, commercial, diplomatic) govern intercountry allocations to asubstantial extent, thereby reducing the margin of flexibility within which aidadministrators must operate.

Furthermore, the Paris Club, which is the chosen instrument of OECD governments tomanage debt problems of less developed countries, adjusted fairly slowly to the specialcircumstances in Africa. The Club was created in 1956 as an ad hoc, informal forum to handleLatin American debt difficulties. Paris Club conditions for rescheduling softened over time,but this process was not based on any clear recognition of the nature of the problem.12 Jaycoxet al. (1986) analyzed the Zambian case to illustrate the functioning of this Club in relationto the rest of the international system. Their overall conclusion was as follows:

Even when the debtor country commits itself to a serious economic reform program(which is approved by the multilateral institutions and accepted by the donor-creditorcommunity), it faces two sorts of problems. First, the process of debt renegotiation is slow,costly in terms of administrative resources, limited in duration, and uncertain as to itslong-term implications. Thus, the debtor has to renegotiate debts not only with bodieslike the Paris Club but also individually with each creditor, a heavy demand on thelimited administrative resources of the debtor government. The rescheduling is valid onlyfor a year, so that the debtor does not receive a clear commitment that its long-termstructural adjustment program will get the debt relief it needs in future years. In addition,since rescheduled debt generally bears commercial rates of interest, the exercise leads toa hardening in the average terms on which external resources are transferred. Suchtemporary relief then leads to a "ballooning" of debt-service payments in a relativelyshort (three to five year) period-too short for most countries in this region [Sub-SaharanAfrica] to achieve a significant economic turnabout.

Second, the existence of a variety of forums that deal in an uncoordinated mannerwith different aspects of debt relief and new credits and grants means that it is practicallyimpossible to achieve the target of transferring a given amount of net resources to thedebtor. There is no institution or forum which currently has overall responsibility forensuring that this sort of target is met. The debtor may be passed back and forth betweenConsultative Groups, the Paris Club, the London Club, and other such bodies without aclear assurance that the resources it needs to carry through its reform program will becommitted. This uncertainty, and the real risk that the resource transfer realized will fallshort of target, acts as a considerable disincentive to persisting in the painful andpolitically costly reform programs (p. 179).

Donors and creditors have been more supportive of SSA reforms in recent years than theywere in the early 1980s. Real net disbursements of official development assistance rose muchmore sharply after 1984 than before, and relatively high priority has been assigned tocountries undertaking reforms (World Bank and UNDP 1989, p. 3). These changes haveenabled reforming countries to reverse the protracted decline in the volume of their imports

12. For a history of the Paris Club and how it reacted gradually to the special circumstances of SSA, seeKlein (1987).


and of investment. Such encouraging signs hold out the hope that it will be possible in thefuture to avoid the heavy social costs of adjustment and to reconcile the goals ofstabilization, structural adjustment, and long-run development. In December 1987 the WorldBank and several bilateral donors launched a Special Program of Assistance for low-incomedebt-distressed countries (World Bank 1988b, p. xxxv). This was followed by a decision inJune 1988 at the Toronto Summit to enable individual creditor governments to choose among amenu of rescheduling options for low-income debtors including partial forgiveness, extensionof maturities, and concessional interest rates. Despite this progress, there is no agreement yeton how to deal with large accumulated arrears in some debtor countries and on how to getcommercial bank creditors to share in the required debt relief action.

The IMF and the World Bank

The IMF was heavily engaged in Sub-Saharan Africa during the 1980s. The number ofstandbys rose very rapidly as did the amount of financial assistance. An increasingproportion of this assistance was supplied on a highly conditional basis-that is, subject toimplementation of specific, quantitatively defined policy measures within an agreed timeframe. The aim was to revolve the IMF's limited resources and to restore financial balancesin member countries as soon as possible. The IMF also played an important role as a "gatekeeper" for other sources of finance who relied on IMF judgments regarding the policyperformance of SSA countries before extending assistance.

The effectiveness of the IMF in Sub-Saharan Africa is a controversial subject on whichthere is a growing literature (Killick 1984; Helleiner 1985, 1986b; Green 1986, Taylor 1988).My analysis earlier in this chapter questions the desirability and feasibility of settingambitious and specific stabilization targets that reflect a short-term orientation. These havebeen the characteristics of IMF standbys. Even agreements under the rubric of the IMF'sExtended Fund Facility (EFF)-which visualized adjustment over the medium term-havehad many of the same features. The general policies of the IMF put the emphasis onavoiding prolonged use of its resources. Therefore, members in balance of payment difficultiesare required to secure a rapid reduction in their deficits largely by curtailing demand. TheIMF has not differentiated enough among member countries and their special problems. Allthe countries have been expected to achieve a quick turnaround and to repay the IMF overthree to five years. Many of these agreements broke down in the course of implementation,and in several cases there was a hiatus between one agreement and its successor.Furthermore, obligations owed to the IMF have become a sizable part of the total debt in anumber of SSA countries. 13 In some cases accumulated arrears now pose formidabledifficulties in continuing the policy effort.

In response to some of these difficulties, the IMF established the Structural AdjustmentFacility (SAF) in March 1986 and the Enhanced Structural Adjustment Facility (ESAF) inJuly 1988. These innovations have increased the Fund's flexibility in dealing with theproblems of low-income countries with limited capacity to adjust over the short term. Loansfrom these facilities carry concessional financial terms. Had these new facilities beencreated in 1980, some of the trauma of recent years might have been avoided. However, asLance Taylor (1988, p. 154) points out, "new ideas enter establishment thinking at bestgradually. Especially in an institution as hermetic as the International Monetary Fund,percolation of alternative ways of viewing the world is bound to be slow."

The World Bank played an important role in Sub-Saharan Africa by financingdevelopment projects. In 1980 it began policy-based lending as a new initiative aimed atstructural adjustment. During the 1980-86 period, the Bank made twenty structural adjustment

13. Debt service owed to the IMF was 29 percent of the total during the 1985-90 period in Zambia, 28percent in Malawi, 18 percent in Zaire, 17 percent in Somalia, and 16 percent in the Sudan (Jaycox et al.1986, p. 52).


loans and twenty-eight sectoral adjustment loans, comprising 16 percent of the total amountlent to Sub-Saharan Africa.14 Its analytical work that focused on policy issues expandedenormously. The policy dialogue between the Bank and member governments changed from"gentle persuasion" (Please 1984, p. 26) to a structured and speedy search for specific policyagreements. Consultative groups of donors chaired by the Bank transformed their meetingsfrom ones at which international finance was mobilized for development projects to sessionsat which SSA governments explained their policy programs and obtained pledges of donorsupport to carry out economic reforms.

When policy-based lending was introduced in 1980, the Bank visualized a series ofoperations spread out over three to five years in individual member countries. It wasexpected that at the end of five to seven years, successful countries would have overcome theexogenous shocks and the internal maladjustments that had precipitated the acutedifficulties. The original policy statement on policy-based lending did not differentiatebetween the problems of Sub-Saharan Africa and those of other regions despite the urging ofsome staff. The Bank's time horizon for policy-based lending was longer than the IMF's, butnot much longer. Recent Bank assessments recognize that in Sub-Saharan Africa (and in othercountries), it will take a fairly long time to cure the disease. Just how long, nobody isprepared to say. Fortunately, the initial miscalculation by the Bank and the currentvagueness do not cause as much of a problem (as in the case of the IMF during the early1980s) because most policy-based loans in Sub-Saharan Africa carry very soft IDA terms.

What has handicapped the Bank in Africa is the inadequate scale of recent IDAreplenishments. In real terms, the amount of IDA lending to Sub-Saharan Africa rose by only6 percent between 1980 and 1986. Even this increase could not have occurred without reducingthe proportion of total IDA funds going to India and China from 45 percent to 23 percent forwhich there was no good justification. To make up for inadequate IDA reserves, a SpecialFacility for Africa was established in 1985 that added about 10 percent to IDA commitments.

Donors have given IDA the mandate to coordinate foreign aid to Sub-Saharan Africa, butdespite its best efforts the outcome (in terms of the quantity, quality, and timeliness ofresource transfers) remains unsatisfactory. Elsewhere I have examined three optionsregarding future improvements in aid policies, including revised criteria for IDA allocation,a new and vigorous approach to aid coordination, and a solution to the "adding up" problemin this context (Gulhati and Nallari 1988). It is important that the international systemadopt a business-like approach to defining the foreign resources required to buttress reformprograms and a fool-proof process for delivering such help.

Both the IMF and the World Bank have determined policy conditionality in Sub-Saharan Africa with the former playing the lead role (Please 1984, p. 79). The Bank hasnot extended policy-based loans to Sub-Saharan Africa without the recipient country havingan IMF agreement signed or in an advanced stage of negotiation (Please 1984, p. 78).Although the two institutions have made substantial efforts to intensify their collaboration,most observers agree that the results remain unsatisfactory (Please 1984, p. 69; Williamson1982, p. 620; Killick 1984, 1987; Helleiner 1986b, p. 60). The IMF and the Bank werecommitted to the same fundamental objectives, but they still disagree with respect to theweights attached to these objectives, the time-horizon for achieving these objectives, thedesirability and feasibility of using various types of policy instruments, and the importanceof institutionalizing the policy function in member countries.

Dissatisfaction with the IMF and World Bank collaboration surfaced openly in 1985 inthe context of the Baker Plan. This plan emphasized "adjustment through growth" insteadof adjustment through demand reduction. Its main focus was on middle-income debtorcountries, but it also had some relevance for low-income Sub-Saharan Africa. The UnitedStates wanted the Bank to play a bigger role in low-income countries than it had in theearly 1980s. Under the newly established SAF and ESAF, the Bank and the IMF had joint

14. The current proportion of SAL/SECAL amounts to total lending was 33 percent.


responsibility for helping members to define a policy framework for a three-year adjustmentprogram. The aim is to broadly define policy objectives and measures that will be adoptedover the whole period and to spell out in detail the next year's program. A financing plan isalso drawn up with the objective of raising resources from all donors and lenders. Since I amfocusing on the period up to 1986, I have not assessed how effective these important newchanges are proving to be.

To what extent has IMF or World Bank policy conditionality been implemented? Afterreviewing the results of many investigations covering the 1960s and 1970s, Killick (1984, p.266) concluded that "the IMF has experienced large difficulties in securing governmentcompliance with a number of its key performance criteria, especially since 1973, with fiscaldifficulties being a major source of non-compliance." Haggard (1986, p. 157) analyzed thirtypolicy agreements under the IMF's Extended Fund Facility and found that "twenty-four wererenegotiated, or had payments interrupted, or were quietly allowed to lapse. Of thesetwenty-four, sixteen were formally cancelled by the IMF, virtually all for noncompliance."Zulu and Nsouli (1985) examined thirty-five IMF programs in Sub-Saharan Africa during1980 and 1981 and found implementation to be poor. Mosley studied eleven countries thatsigned SAL agreements with the World Bank. Four of these countries complied with policyconditions to the extent of 90 percent or more, in another four compliance varied from 60 to 90percent, and the remaining three complied less than 60 percent (Mosley 1987, p. 9). A WorldBank study (1988, p. 7) of fifty-one SALs and SECALSs in fifteen countries found that 60percent of all conditions were implemented fully during the loan disbursement period. Theextent of full implementation was relatively high in some policy areas (energy, finance,exchange rate) and relatively low in others (fiscal matters, industry, trade). A study byNelson (forthcoming) concluded that international agencies did not exercise a major impacton the policy histories of thirteen countries.

Many factors explain this mixed record of conditionality. To begin with, the amount ofgenuine consensus between the lenders and the borrowers varies from country to country andeven from one conditional loan to the next in the same country. Even if one assumes fullconsensus, compliance with the loan agreement may be undermined by unexpecteddevelopments (including unforeseen administrative obstacles) during the period ofimplementation. Next, consensus in the borrower country is very difficult to ascertain, giventhe subtleties of interministerial relations, competing party factions, and conflicting economicinterests among pressure groups. Even well-meaning government spokespersons can readilymisjudge the extent to which their colleagues in government or private parties are commnittedto the reforms. Consequently, policy agreements signed in good faith will often unravelduring implementation.

A substantial number of conditional loans signed in the 1980s were not based on genuineagreements. Borrower governments often signed loan documents even though they remainedskeptical about the desirability or feasibility of the policy conditions. The desperate needfor cash compelled them to overlook or minimize their reservations regarding theconditionality. These reservations concerned both the substance of what the lending agencieswanted done and their negotiating style. These sentiments lie behind the following quotefrom Helleiner (1986a, p. 8). His comments are in reference to a seminar held in Nairobi inMay 1985 attended by African governors of the IMF and senior Fund officials, but they alsoare applicable to the World Bank.

There was a remarkable degree of agreement among Governors that Fund staff,missions, and management have not always responded to their problems with theunderstanding that they believe they have a right to expect. Many, including some withreasonably successful Fund programs, suggested that the Fund staff is inadequatelyinformed or insensitive with respect to local conditions and objectives, patronizing in theirrelationships with local professionals, and rigid or powerless or both in their negotiationswith African governments. Whether or not these perceptions are accurate, theirpersistence must be a matter for profound regret.


These perceptions of Africans raise important questions not only about the working styleof IMF and Bank staff but also about their approach to conditionality, which is whatinterests me. Critics claim that while a part of the World Bank's conditionality is beyondcontroversy, there are other aspects (role of the state, the degree of outward orientation, thedistribution of income) on which development theory provides little guidance (Helleiner1986b, p. 52-59; Lewis, 1986, p. 9). Accordingly, the Bank has been criticized for championinga new orthodoxy.1 5 Please (1984, p. 83) reports on widespread perceptions that the Bankpermits "the objectivity of its economic analysis and its criteria for evaluating developmentpolicies to be influenced by the current ideological preferences of the major shareholders-most obviously the ideological preference of the United States in recent years for privateenterprise activity and for the 'magic of the market place."' The attention of theinternational agencies is focused on obtaining agreement on specific policy measures at apoint in time. This orientation reflected the basically optimistic view that adjustmentproblems could be resolved fairly quickly. In the hurry to get things done speedily, theinternational agencies took upon themselves much of the diagnostic work, the definition ofpolicy options, and even the drafting of policy agreements. They gave very little attentionto the determinants of economic policy (historical and political) in individual countries andto systemic issues bearing on the economic policy process in these countries. No matter howwrong these perceptions are, they undermine to some extent the effectiveness of the Bank andthe IMF in bringing about sustained policy reforms.

Looking to the future, international agencies need to acquire a long-term approach topolicy development in Sub-Saharan Africa. This means focusing on the aim of sustainedpolicy change and not on individual policy loans. In this context the international agencieswill have to help governments to acquire the expertise and the confidence needed to takeprimary responsibility for designing and implementing economic policy. The new policieshave to be homespun products adapted to the political and socioeconomic environment ofeach country.

One of the major lessons of the 1980s is that policy-based lending cannot succeed on thebasis of leverage alone. Leverage is the capacity of international organizations to imposetheir policy position on the member government by virtue of their own "financial muscle"and their influence on other sources of finance. International organizations enjoy suchleverage on an episodic basis; it waxes and wanes in response to the financial status of clientcountries. Policies adopted by SSA governments as a result of pure dialogue have a muchbetter chance of being implemented and sustained than policies that are accepted onlybecause of the financial power of the IMF and the Bank. The purpose of such dialogue is toclarify the diagnosis of the economic situation and to improve understanding of the policyposture of each party (Cassen et al. 1986, p. 70). Good dialogue is a two-way interaction inwhich the member government listens to the views of the international organizations andvice versa. Agreement is sought on numerous issues: the functioning of the economy; theweights that should be attached to different policy objectives; the costs and benefits of usingalternative policy instruments; and the timing, scope, speed, and phasing of new policies. Ifagreements are based on good dialogue, government policymakers will see the agreements asa way of pursuing their personal, corporate, or national interest. By contrast, IMF or Bankpolicies accepted under pressure and despite important intellectual reservations are likely to

15. Helleiner (1986b, p. 54) says "the World Bank possesses a fairly consistent approach to policy-based lending. Like the IMF, it stresses monetary and fiscal orthodoxy, appropriate real exchangerates, positive real interest rates, avoidance of administrative controls on external transactions. As far aslonger-term development strategy is concerned, the Bank urges expansion and overall outwardorientation as against import substitution; liberalization of import barriers and movement towardunified import incentives; and maximum reliance upon markets rather than government ownership ordirection in the domestic economy. The Bank's primary emphasis is on price incentives and 'getting theprices right,' and its obvious presumption is that, even in a world of pervasive imperfections, marketscan normally be trusted to achieve that objective better than governments. Even in the world of thesecond-best, its approach is consistently to liberalize that which can be liberalized."


remain foreign "impositions" that the government may resist during implementation or assoon as the pressure eases.

As noted earlier, significant and sustained changes in the policy process in Sub-SaharanAfrica will require basic changes in the attitudes of the top African leaders. It may not beeasy for the international agencies to directly influence these leaders, but they should try.Too much of the dialogue between these agencies and SSA governments involves middle levelBank-IMF staff and civil servants; too little interaction takes place between senior managersof these agencies and top SSA leaders. Everyone recognizes the importance of politicalissues, but they tend to be put under the rug. The few opportunities for a meaningful exchangeat top levels are wasted.

In the future both sides should go beyond the technical issues to discuss more fully thepolitical risks inherent in economic reforms. Learning more about how SSA leaders perceivethese risks can be salutary for top international officials. The latter in turn can draw ontheir experience in other LDCs to convey to top African leaders how such risks werecontained in comparable situations. In view of allegations that international agencies havebecome too hegemonic and doctrinaire in their advocacy of free-market principles, they muststrive to demonstrate their commitment to an open, pluralistic, and pragmatic approach.

The dialogue between SSA civil servants and the staffs of international organizationsalso needs revamping. Analytical and policy studies should be conducted in a more collegiatemanner. The discussions of these studies should be much broader than they have been. Allrelevant policy options and their consequences should be examined, and African officialsfrom all relevant government agencies and other policy actors in member countries shouldparticipate. These changes have been introduced, in principle, in the World Bank, but theirimplementation varies a great deal. More attention should be given to the budgetaryimplications of these changes. A participative style for report writing and a program of fulldiscussion will require considerably more time and resources than are typically availablenow.

Conclusion

While policy change everywhere is a risky endeavor, it is particularly risky in manySSA countries. First, the immediate beneficiaries of the new policies-that is, thepeasants-are politically weak, and the development coalition that stands behind policieserected during the 1970s is relatively strong. The peasants in this region are numerous,geographically scattered, and ethnically divided Second, the transitional period betweenthe initiation of reforms and the point at which their impact is fully registered on theeconomy is likely to be a long one, given the stage of underdevelopment. Third, the corpus ofsocial theory that underpins the design of reforms contains many gaps, especially in thecontext of Sub-Saharan Africa. The effects of the new policies on the economy remain veryuncertain. Fourth, many SSA countries are especially vulnerable to exogenous shocks that canundermine the favorable results of reforms. Political violence, weather vicissitudes,international price fluctuations and, ill-considered actions of foreign actors can be verydisruptive.

In the sample of twelve countries, many factors impeded intensive and sustained economicreforms between 1980 and 1986, but perhaps the crucial one was the short time horizon ofpolitical leaders. They lacked a strategic vision and relied on patron-client relations tosurvive. Many of the leaders preferred to "coast along," avoiding the risks that basicreforms would entail. What factors make for good leaders is an unanswered question, butsome combination of adversity and new ideas (self learning, mutual learning) can stimulatefavorable changes. Once the top leadership has become reform-minded, the process ofintroducing and sustaining new economic policies and supportive institutional changes cantake hold. The new orientation of leaders leads to a reevaluation; the dangers of "coastingalong" now appear to be bigger than the risks associated with basic reforms. New political


coalitions have to be built to support reforms. This entrepreneurial statecraft requires avigorous public education program as well as a policy design that takes fully into account notonly technical economic factors but also ways to elicit the cooperation of all parties whohave the power to block reforms. The timing, scope, and sequence of policy packages must beresponsive to these political parameters.

In this interdependent world some of the risks of policy change arise from the behavior offoreign actors who have a decisive role in determining the terms of trade, capital flows, andthe interpretation of conditionality. Stabilization and structural adjustment in Sub-SaharanAfrica would be much easier to achieve if the world economy were characterized bysustained high growth and if the policies of major powers in the areas of trade and financewere supportive. The IMF and the World Bank which have played and undoubtedly willcontinue to play a key role. It will be very important for these multilaterals to have enoughconcessional finance and to deal flexibly with the distinctive problems of individual membercountries. General rules about the need to revolve funds quickly-rules that in the pastproduced a pronounced short-term policy conditionality-need to be avoided. Also to beeschewed is the perception that these organizations are promoting a new policy orthodoxy.The IMF and the Bank must help Sub-Saharan Africa governments increase indigenousparticipation in the preparatory phase, the decisionmaking phase, and the implementationphase of economic reform.

How should the controversies surrounding policy reform in Sub-Saharan Africa bereassessed in the context of this analysis? I will examine again, as I did in Chapter 1, theviews of Glickman (1988) and the Economic Commission for Africa (ECA undated). Glickmanvoiced the sentiment of many political scientists when he implied that the economic reformsemphasizing austerity, urged by foreign donors, could undermine the African state. Such aview has a measure of validity. Authoritarian rule in many SSA countries depends onpatron-client relations, and a very sharp reduction in public expenditure would weaken SSAgovernments' capacity to buy support from various groups. This view also highlights thedanger of reforms imposed by outsiders for which there is little domestic support. Glickman'sthesis, however, appears to be derived from a very static view of political legitimacy andfrom a misleading presumption that macro policy managers in Sub-Saharan Africa do nothave an interest in restoring financial balance. Experience over the 1980-86 perioddemonstrates that reform involving substantial austerity can be adopted and sustained,provided it is well designed and provided political leaders recognize the need for suchreform.

I agree with three conclusions in the recent ECA report. First, economic policy has becomethe prisoner of short-term crisis management and the "siege mentality" (ECA undated, p. 33).Long-run aspects have been neglected. I agree that the aim should be not only to adjust toshocks but also to solve structural problems. "Adjustment with transformation" is a muchbetter goal than just correcting financial imbalances. Second, economic policy has been on a"short leash" (p. 8). By this characterization, the ECA is highlighting the coercive orleverage aspects of conditionality that can be counterproductive. Third, the report is at onewith orthodox approaches in insisting on the need to improve financial management andaccountability, to increase parastatal efficiency, to improve agricultural incentives, topromote the diversification of exports, and to restrain defense and nonproductive governmentoutlays. While accepting the need to adjust, the ECA concludes that many policy packagesduring the 1980s overemphasized the objective of reaching a financial balance quickly andthat a preoccupation with curtailing aggregate demand led to the neglect of measures aimedat raising production. These three conclusions of ECA tie in with my own findings in thisstudy.

Other parts of the ECA report, however, are far from convincing. The report advocatesnumerous administrative controls on prices, imports, and credit, as well as complex regimes ofdifferential interest rates and multiple exchange rates. The need for "serious and goodgovernment" to administer such policies is acknowledged (p. 45). It is also recognized thatmany of the current regimes in Sub-Saharan Africa do not meet this standard. The ECA


report goes on to demand "a reorganization and/or realignment of the decision-makingprocess" (p. 49). It prescribes "democratization of the decision-making process at national,local, and grassroots levels" (p. 50). There is no discussion, however, of what forces willbring about such a radical political change or of the various stages in the transition betweenthe authoritarian regimes of today and the emergence of democracy. In the absence of such arigorous analysis, the ECA prescription strikes an unrealistic note. States need to intervene inthe economy to correct market failures, but they can do so effectively only if theythemselves are robust instruments of social welfare. Personal rule and the dominance ofpatron-client relations in Africa today are unlikely to disappear in a hurry. History suggeststhat such changes occur over decades if not generations. Meanwhile, it is important to assesswhat reforms are feasible in the context of the existing stage of political development.

ECA's critique of what it calls "orthodox adjustment policies" of the IMF and the WorldBank has to be seen as an attempt to redress the excesses of the antistate counterrevolution indevelopment philosophy that occurred in the late 1970s and 1980s. Such a redressal isrequired and is under way to some extent. ECA's analysis, however, is not persuasive. Forexample, ECA is categorical in denouncing generalized devaluation and import liberalization(pp. 37, 38), while my own assessment of these policies is much less sweeping. Anotherexample is the ECA statement that "reductions in budget deficits must not be accomplishedat the expense of expenditures on the social sector, i.e., education, health, and other socialinfrastructure" (pp. 33, 34). Such an unqualified exclusion of all such outlays from budget cutsis difficult to countenance, even though I agree that such outlays typically have highpriority and need a protective umbrella. There is plenty of scope for improving costeffectiveness, increasing cost recovery, and restructuring public expenditures on education andhealth, for instance. While the ECA is critical of the mechanistic aspects of orthodoxpolicies, it does not hesitate to advance mechanistic rules of its own-for example, 30 percentof government outlays should go to the social sector (p. 34), and 20 to 25 percent of publicinvestment should be allocated to agriculture. It is not clear how these rules are consistentwith the report's emphasis on flexibility in designing policy packages and on findingsolutions tailored to the specific circumstances of each country.

REFERENCES

Abernethy, David B. 1988. "Bureaucratic Growth and Economic Stagnation in Sub-SaharanAfrica." In Stephen K. Commins, ed., Africa's Development Challenges and theWorld Bank. Boulder, Colo.: Lynne Rienner.

Adedeji, Adebayo. 1986. "Administrative Adjustments and Responses to Changes in theEconomic Environment." In African Association for Public Administration andManagement, The Ecology of Public Administration and Management in Africa. NewDelhi: Vikas Publishing House Pvt. Ltd.

Ake, Claude. 1981. The Political Economy of Africa. Essex, England: Longman.

Allison, Graham. 1971. Essence of Decision: Explaining the Cuban Missile Crisis. Boston:Little, Brown.

Alverson, Hoyt. 1978. Mind in Heart of Darkness. New Haven, Conn.: Yale University Press.

African Debt and Financing. 1986. Institute for International Economics Special Report 5.Washington, D.C.: Institute for International Economics.

Balassa, Bela. 1984. "Adjustment Policies and Development Strategies in Sub-SaharanAfrica, 1973-78." In Moshe Syrquin, Lance Taylor, and Larry E. Westphal, eds.,Economic Structure and Performance. Essays in honor of Hollis B. Chenery. NewYork: Academic Press.

Bates, Robert H. 1981. Markets and States in Tropical Africa: The Political Basis ofAgricultural Policies. Berkeley: University of California Press, 1981.

. 1983. Essays on the Political Economy of Rural Africa. New York: CambridgeUniversity Press.

_____. 1988. "The Political Basis for Agricultural Policy Reform." In Stephen K.Commins, ed., Africa's Development Challenges and the World Bank. Boulder,Colo.: Lynne Rienner.

Bauer, Peter T. 1972. Dissent on Development. London: Widenfeld & Nicolson.

Baum, Warren C., and Stokes M. Tolbert. 1985. Investing in Development: Lessons of WorldBank Experience. New York: Oxford University Press.

Berg, Elliot. 1987. "Obstacles to Liberalizing Agricultural Markets in Developing Countries."In Dieter Elz, ed., Agricultural Marketing Strategy and Pricing Policy. Washington,D.C.: World Bank.

Berg, Elliot, and Mary M. Shirley. 1987. Divestiture in Developing Countries. World BankDiscussion Paper 11. Washington, D.C.: World Bank.

101


Bgoya, Walter, and Goran Hyden. 1987. "The State and the Crisis in Africa: In Search of aSecond Liberation." Development Dialogue, vol. 2.

Bhagwati, Jagdish N., and T. N. Srinivasan. 1975. Foreign Trade Regimes and EconomicDevelopment: India. New York: Columbia University Press.

Brown, Richard. 1986. "International Responses to Sudan's Econornic Crisis: 1978 to the April1985 Coup d'Etat." Development and Change, vol. 17

_ . 1988. "A Background Note on the Final Round of Economic Austerity MeasuresImposed by the Nimeiry Regime: June 1984-March 1985." In Tony Barnett and AbbasAbdel Karim, eds., Sudan: State, Capital and Transformation. New York: CroomHelm.

Buchanan, James M., and Gordon Tullock. 1962. The Calculus of Consent: Logical Foundationsof Constitutional Democracy. Ann Arbor, Mich.: University of Michigan Press.

Callaghy, Thomas M. 1986a. "The Political Economy of African Debt: The Case of Zaire."In John Ravenhill, ed., Africa in Economic Crisis. New York: Columbia UniversityPress.

. 1986b. "The State and the Development of Capitalism in Africa: Theoretical,Historical, and Comparative Reflections." In Donald Rothchild and NaomiChazan, eds., The Precarious Balance: State and Society in Africa. Boulder, Colo.:Westview Press.

Cartwright, John R. 1983. Political Leadership in Africa. New York: St. Martin's Press.

Cassen, Robert, et al. 1986. Does Aid Work? Oxford: Claredon Press.

Chhibber, Ajay. 1988. "Raising Agricultural Output: Price and Nonprice Factors." Financeand Development, vol. 25.

Cleaver, Kevin M. 1985. The Impact of Price and Exchange Rate Policies on Agriculture inSub-Saharan Africa. World Bank Staff Working Paper 728. Washington, D.C.:World Bank.

Cleaves. P. 1980. "Implementation Amnidst Scarcity and Apathy: Political Power and PolicyDesign." In Merilee S. Grindle, eds., Politics and Policy Implementation in theThird World. Princeton, N.J.: Princeton University Press.

Covell, Maureen. 1987. Madagascar: Politics, Economics and Society. New York: FrancesPinter.

Dror, Yehezbel. 1986. Policy-Making Under Adversity. New Brunswick, N.J.: TransactionBooks.

ECA (Economic Commission for Africa). undated. African Alternative Framework toStructural Adjustment Programs for Socio-Economic Recovery and Transformation.E/ECA/C.M.15/6/Rev 3.

References 103

Eicher, Carol K., and Doyle Baker. 1982. Research on Agricultural Development in Sub-Saharan Africa: A Critical Survey. International Development Paper 1. Lansing,Mich.: Michigan State University.

Ekeh, P. 1975. "Colonialism and the Two Publics in Africa: A Theoretical Statement."Comparative Studies in Society and History, vol. 17, no. 1.

Elliot, Charles. 1988. "Structural Adjustment in the Longer Run: Some UncomfortableQuestions." In Stephen K. Commins, ed., Africa's Development Challenges and theWorld Bank. Boulder, Colo.: Lynne Rienner.

Etzioni, A. 1985. "Making Policy for Complex Systems: A Medical Model for Economics."Journal of Policy Analysis and Management, vol. 4, no. 3.

Fearon, James D. 1988. "International Financial Institutions and Economic Policy Reform inSub-Saharan Africa." The Journal of Modern African Studies, vol. 26, no. 1.

Fieldhouse, D.K. 1986. Black Africa, 1945-80: Economic Decolonization and ArrestedDevelopment. London: Allen & Unwin.

Friedman, Milton. 1970. The Counter-Revolution in Monetary Theory. London: Institute ofEconomic Affairs.

Ghosh, Jayati. 1986. "Foreign Debt and Economic Development: The Case of Zaire."Development and Change, vol. 17.

Glickman, Harvey, ed. 1988. The Crisis and Challenge of African Development.Contributions in Afro-American Studies, 112. Westport, Conn.: Greenwood Press.

Godfrey, Martin. 1987. "Stabilization and Structural Adjustment of the Kenyan Economy,1975-85: An Assessment of Performance." Development and Change, vol. 18.

Gordon, David. 1983. Which Way for Zimbabwe? Development Dilemmas and Prospects.Discussion Paper 104. Ann Arbor, Mich.: University of Michigan, Center forResearch on Economic Development.

Green, Reginald. 1986. The IMF and Stabilization in Sub-Saharan Africa: A Critical Review.IDS Discussion Paper 216. Brighton, England: Institute of Development Studies,University of Sussex.

. 1987. "Articulating IMF Stabilization Programmes with Structural Adjustment inSSA Agriculture." Paper delivered at ODI conference on the "Design and Impact ofAdjustment Programmes on Agriculture and Agricultural Institutions." London:Overseas Development Institute, Regent's College.

Grindle, Merilee S., ed. 1980. Politics and Policy Implementation in the Third World.Princeton, N.J.: Princeton University Press.

Gulhati, Ravi. 1988. Political Economy of Reform in Sub-Saharan Africa: Report of theWorkshops on the Political Economy of Structural Adjustment and theSustainability of Reform. EDI Policy Seminar Report 8. Washington, D.C.: WorldBank.


. 1989a. Impasse in Zambia: The Economics and Politics of Policy Reform. EDIDevelopment Policy Case Studies no. 2. Washington, D.C.: World Bank.

. 1989b. Malawi: Promising Reforms, Bad Luck. EDI Development Policy CaseStudies no. 3. Washington, D.C.: World Bank.

. 1990. "Emerging Trade Reforms in Sub-Saharan Africa: How Have They AffectedManufacturing?" In H. Singer and S. Sharma, eds., Growth and External DebtManagement. United Kingdom: McMillan Press.

Gulhati, Ravi, and Gautam Datta. 1983. Capital Accumulation in Eastern and SouthernAfrica: A Decade of Setbacks. World Bank Staff Working Paper 562. Washington,D.C.: World Bank.

Gulhati, Ravi, and Raj Nallari. 1988. "Reform of Foreign Aid Policies: The Issue of Inter-Country Allocation in Africa." World Development, vol. 16, no. 10.

. forthcoming. Mauritius: A Case of Successful Stabilization and Recovery. EDIDevelopment Policy Case Studies. Washington, D.C.: World Bank.

Gulhati, Ravi, and Satya Yalamanchili. 1988. "Contemporary Policy Responses to EconomicDecline in Africa." In Harvey Glickman, ed., The Crisis and Challenge of AfricanDevelopment. Contributions in Afro-American Studies, 112. Westport, Conn.:Greenwood Press.

Gulhati, Ravi, Swadesh Bose, and Vimal Atukorala. 1985. Exchange Rate Policies in Easternand Southern Africa, 1965-1983. World Bank Staff Working Paper 720. Washington,D.C.: World Bank.

. 1986. "Exchange Rate Policies in Africa: How Valid is the Skepticism?"Development and Change, vol. 17.

Gulhati, Ravi, and Uday Sekhar. 1981. Industrial Strategy for Late Starters: The Experienceof Kenya, Tanzania, and Zambia. World Bank Staff Working Paper 457.Washington, D.C.: World Bank.

Gulhati, Ravi, and Vimal Atukorala. 1985. "Import Instability and External Reserves inEastern and Southern Africa." Washington, D.C.: World Bank. Processed.

Hadjor, Kofi B. 1987. On Transforming Africa: Discourse with Africa's Leaders. Trenton, N.J.:Africa World.

Haggard, Stephan, and Robert Kaufman. 1989. "The Politics of Stabilization and StructuralAdjustment." In Jeffrey D. Sachs, ed., Developing Country Debt and EconomicPerformance, vol. 1. Chicago: University of Chicago Press.

Hampson, Fen 0. 1986. Forming Economic Policy: The Case of Energy in Canada and Mexico.New York: St. Martin's Press.

Hartmann, Jeannette. 1988. "Crisis and Reforms." In P. Fruhling, ed., Recovery in Africa.Stockholm: Swedish Ministry of Foreign Affairs.

References 105

Harvey, Charles. 1987. Successful Macroeconomic Adjustment in Three Developing Countries:Botswana, Malawi, and Papua New Guinea. EDI Development Policy Case Series.Washington, D.C.: World Bank.

Havnevik, Kjell J., ed. 1987. The IMF and the World Bank in Africa: Conditionality, Impactand Alternatives. Uppsala, Sweden: Scandinavian Institute of African Studies.

Hazelwood, Arthur. 1975. Economic Integration: The East African Experience. Exeter, N.H.:Heinemann Educational Books.

Helleiner, Gerald K. 1985. "Aid and Liquidity: The Neglect of Sub-Saharan Africa andOthers of the Poorest in the Emerging International Monetary System." Journal ofDevelopment Planning, no. 15.

. 1986a. Africa and the International Monetary Fund. Washington, D.C.:International Monetary Fund.

_ . 1986b. "Policy-Based Program Lending: A Look at the Bank's New Role." InRichard E. Feinberg et al., Between Two Worlds: The World Bank's Next Decade.U.S.-Third World Policy Perspectives 7, Overseas Development Council. NewBrunswick, N.J.: Transaction Books.

_=____. 1986c. "The Question of Conditionality." In Carol Lancaster and John Williamson,eds. African Debt and Financing. Institute for International Economics SpecialReport 5. Washington, D.C. : Institute for International Economics.

Heller, Peter, and Alan Tait. 1983. Government Employment and Pay: Some InternationalComparisons. IMF Occasional Paper 24. Washington, D.C.: International MonetaryFund.

Hirschman, Albert. 1981. Essays in Trespassing: Economics to Politics and Beyond. New York:Cambridge University Press.

Hodd, Michael, 1987. "Africa, the IMF and the World Bank." African Affairs, vol. 86, no.344.

Hyden, Goren. 1983. No Shortcuts to Progress: African Development Management inPerspective. Berkeley: University of California Press.

. 1988. "State and Nation Under Stress." In Recovery in Africa. Stockholm:Swedish Ministry of Foreign Affairs.

ILO (International Labor Organization). 1985. Government Regulation of Wages andIndustrial Relations in the Private Sector. Geneva.

_ . 1986. Youth, Employment and Employment Programs in Africa. Addis Ababa:International Labor Organization, Jobs and Skills Programme for Africa.

Institute of Development Studies. 1983. "What Agenda for Action?" IDS Bulletin, vol. 14, no.2, January.

Jackson, P.M. 1983. The Political Economy of Bureaucracy. New York: Barnes & Noble Books.


Jackson, Robert H., and Carl G. Rosberg. 1982. Personal Rule in Black Africa: Prince,Autocrat, Prophet, Tyrant. Berkeley: University of California Press.

. 1986. "Sovereignty and Underdevelopment: Juridical Statehood in the AfricanCrisis." The Journal of Modern African Studies.

Jamal, Vali. 1988. "Somalia: Understanding an Unconventional Economy." Development andChange, vol. 19.

Jansen, Doris J. 1988. Trade, Exchange Rates, and Agricultural Pricing Policy in Zambia: ThePolitical Economy of Agricultural Pricing Policy. A World Bank ComparativeStudy. Washington, D.C.: World Bank.

Jaycox, Edward V.K., et al. 1986. "The Nature of the Debt Problem in Eastern and SouthernAfrica." In Carol Lancaster and John Williamson, eds., African Debt and Financing.Special Report No. 5. Washington, D.C.: Institute for International Economics.

Johnson, Harry, G. 1971. "The Keynesian Revolution and the Monetarist Counter-Revolution."American Economic Review, vol. 61.

Kadhani, Xavier. 1986. "The Economy: Issues, Problems and Prospects." In Ibbo Mandana, ed.,Zimbabwe: The Political Economy of Transition, 1980-1986. London: Codesria Books.

Kamuntu, Ephraim. 1987. Implementation Issues in African Adjustment: The Case of Uganda.unpublished.

Kerr, T.C. 1985. "Trends in Agricultural Price Protection." World Development Report, 1986.Background Paper 11. New York: Oxford University Press.

Keynes, John M. 1936. The General Theory of Employment, Interest and Money. London:MacMillan.

Killick, Tony. 1983. "Kenya, the IMF and the Unsuccessful Quest for Stabilization." In JohnWilliamson, ed., IMF Conditionality. Washington, D.C.: Institute for InternationalEconomics.

_. 1984. The Quest for Economic Stabilization: The IMF and the Third World.London: Heinemann.

. 1987. "Reflections on the IMF/World Bank Relationship." In Kjell J. Havnevik,ed., The IMF and the World Bank in Africa: Conditionality, Impact andAlternatives. Uppsala, Sweden: Scandinavian Institute of African Studies.

1989. A Reaction Too Far: Economic Theory and the Role of the State inDeveloping Countries. London: Overseas Development Institute.

Klein, Thomas M. 1987. "Debt Relief for African Countries: A New Approach by OfficialCreditors in the Paris Club." Finance and Development, vol. 24, no. 4.

Krueger, Ann 0. 1974. "The Political Economy of the Rent-Seeking Society." AmericanEconomic Review, vol. 64.

References 107

Krueger, Ann O., Maurice M. Schiff, and Alberto Valdes. 1988. "Agricultural Incentives inDeveloping Countries: Measuring the Effect of Sectoral and Economywide Policies."The World Bank Economic Review, vol. 2, no. 3.

Lancaster, Carol. 1988. Political Economists and Policy Reformers in Africa. EDI WorkingPaper. Washington, D.C.: World Bank.

Legum, Colin, ed. 1986. African Contemporary Record, 1984-85, vol. 17.

Le-e, Uma. 1986. "Comparative Advantage and Structural Transformation: A Review ofAfrica's Economic Development Experience." World Bank, Development ResearchDepartment. Washington, D.C. Processed.

. 1987. "Agriculture and Infrastructure." Paper prepared for a symposium onTransportation and Structural Adjustment organized by the World BankTransportation Department. Washington, D.C.: World Bank.

. 1988. Price and Nonprice Aspects of Sources of Growth in East AfricanAgricultures: Implications for Governments and Donors. Washington, D.C.: WorldBank.

Lele, Uma, and Robert E. Christiansen. 1988. "Markets, Marketing Boards and Cooperatives:Their Role in Agricultural Development." Paper prepared for a conference on"Agricultural Growth and Market Town Development" in Lilongwe, Malawi, May.

Lewis, John P. 1986. "Development Promotion: A Time for Regrouping." In John P. Lewis andValeriana Kallab, eds., Development Strategies Reconsidered. OverseasDevelopment Council, U.S.-Third World Policy Perspectives 5. New Brunswick, N.J.:Transaction Books.

Lewis, John P., and Valeriana Kallab, eds. 1986. Development Strategies Reconsidered.Overseas Development Council, U.S.-Third World Policy Perspectives 5. NewBrunswick, N.J.: Transaction Books.

Lindauer, David. 1988. Government Pay and Employment Policies and GovernmentPerformance in Developing Economies. Policy, Planning, and Research WorkingPaper Series, 42. Washington, D.C.: World Bank.

Lindauer, David, Oey Messook, and Parita Suebsaeng. 1988. "Government Wage Policy inAfrica: Some Findings and Policy Issues." The World Bank Research Observer, vol.3, no. 1.

Lungu, G. 1983. "Administrative Responsibility in a Developing Country: TheoreticalConsiderations and the Case of Zambia." Public Administration and Development,vol. 3.

Mandaza, Ibbo. 1986. "The Political Economy of Transition." In Ibbo Mandaza, ed.,Zimbabwe: The Political Economy of Transition, 1980-1986. London: Codesria Books.

Marenin, Otwin. 1987. "The Managerial State in Africa: A Conflict Coalition Perspective."In Zaki Ergas, ed., The African State in Transition. New York: St. Martin's Press.

Mauritius, Government of. 1984. Industrial Policy Review. Study by the Center forDevelopment Technology for the Government of Mauritius. Port Louis.


Montgomery, John. 1986a. "Life at the Apex: The Functions of Permanent Secretaries in NineSouthern African Countries." Public Administration and Development, no. 6.

. 1986b. "Bureaucratic Politics in Southern Africa." Public Administration Review,September-October.

. 1987. "How African Managers Serve Development Goals." Comparative Politics,April.

M )sley, Paul. 1986. "The Politics of Economic Liberalization: USAID and the World Bank inKenya, 1980-84." African Affairs, vol. 85, no. 338.

_. 1987. Conditionality as Bargaining Process: Structural-Adjustment Lending.International Finance Series, 168. Princeton, N.J.: Princeton University, InternationalFinance Section, Department of Economics.

Ndulu, Benno J. 1986. "Governance and Economic Management." In Robert J. Berg and JenniferSeymour Whitaker, eds., Strategies for African Development. Berkeley: Universityof California Press.

. -1987. Stabilization and Adjustment Policies and Programmes: Tanzania. CountryStudy 17. Helsinki: World Institute for Development Economics Research of theUnited Nations University.

Nellis, John R. 1986. Public Enterprises in Sub-Saharan Africa. World Bank Discussion Paper1. Washington, D.C.: World Bank.

. 1988. Contract Plans and Public Enterprise Performance. Policy, Planning, andResearch Working Paper 118. Washington, D.C.: World Bank.

Nelson, Joan M. 1986. "The Diplomacy of Policy-Based Lending." In Richard E. Feinberg etal., Between Two Worlds: The World Bank's Next Decade. Overseas DevelopmentCouncil, U.S.-Third World Policy Perspectives 7. New Brunswick, N.J.: TransactionBooks.

, ed. forthcoming. Economic Crisis and Policy Choice: The Politics of Adjustment inthe Third World. Princeton, N. J.: Princeton University Press.

Nelson, Joan, et al. 1989. The Politics of Economic Adjustment: Fragile Coalitions. NewBrunswick, N.J.: Transaction Books.

Nzongola-Ntalaja. 1986. "Crisis and Change in Zaire, 1960-1985." In Nzongola-Ntalaja, ed.,The Crisis in Zaire: Myths and Realities. Trenton, N.J.: Africa World.

Ojuka, Aloo. Undated. "Politics and Development in Africa." In A. Ojuka and W. Ochieng',eds., Politics and Leadership in Africa. Kenya: East African Literature Bureau.

O'Neill, Norman. 1988. "Class and Politics in the Modern History of Sudan." In NormanO'Neill and Jay O'Brien, Economy and Class in Sudan. Brookfield, Vt.: GowerPublishing Co.

Ottaway, Marina. "Post Numeiri Sudan: One Year on." Third World Quarterly. July.

Payer, Cheryl. 1983. "And the World Bank.' Third World Quarterly, vol. 5, no. 4.

References 109

Picard, Louis. 1987. The Politics of Development in Botswana: A Model for Success? Boulder,Colo.: Lynne Rienner.

Pinto, Brian. 1988. Black Market Premia, Exchange Rate Unification, and Inflation in Sub-Saharan Africa. Policy, Planning, and Research Working Paper 37. Washington,D.C.: World Bank.

Please, Stanley. 1984. The Hobbled Giant: Essays on the World Bank. Boulder, Colo.:Westview Press.

Pratt, R.C. 1979. "Reflections of a Democratic Socialist." In Bismarck Mwansau and R.C.Pratt, eds., Toward Socialism in Tanzania. Toronto: University of Toronto Press.

Prebisch, Raul. 1959. "Commercial Policy in the Underdeveloped Countries." AmericanEconomic Review, vol. 49.

Raphaeli, Nimrod, Jacques Roumani, and A.C. Mackellar. 1985. Public Sector Management inBotswana: Lessons in Pragmatism. World Bank Staff Working Paper 709.Washington, D.C.: World Bank.

Rothchild, Donald, and Robert L. Curry. 1978. Scarcity, Choice, and Public Policy in MiddleAfrica. Berkeley: University of California Press.

Sandbrook, Richard. 1986. "The State and Economic Stagnation in Tropical Africa." WorldDevelopment, vol. 14, no. 3.

Schwab, Peter. 1985. Ethiopia: Politics, Economics, Society. London: Frances Pinter.

Shivji, I.G. 1976. Class Struggles in Tanzania. New York: Monthly Review.

Shuttleworth, Graham. 1989. "Policies in Transition: Lessons from Madagascar." WorldDevelopment, vol. 17, no. 3.

Singh, Ajit. 1986. "And the IMF: The Analytics of Alternative Adjustment Programmes."Development and Change, vol. 17.

Sithole, Masipula. 1986. "The General Elections: 1979-1985." In Ibbo Mandaza, ed.,Zimbabwe: The Political Economy of Transition, 1980-1986. London: Codesria Books.

Stark, Frank M. 1986. "Theories of Contemporary State Formation in Africa: AReassessment." The Journal of Modern African Studies, vol. 24, no. 2.

Streeten, Paul. 1987. "Structural Adjustment: A Survey of the Issues and Options." WorldDevelopment, vol. 15, no. 12.

Taylor, Lance. 1988. Varieties of Stabilization Experience: Towards Sensible Macroeconomicsin the Third World. New York: Oxford University Press.

Tinbergen, Jan. 1967. Development Planning. London: Weidenfeld & Nicholson.

Toye, John. 1987. Dilemmas of Development: Reflections on the Counter-Revolution inDevelopment Theory and Policy. New York: Basil Blackwell.


Uganda, Republic of. 1982. Report of the Public Service Salaries Review Commission.Kampala: Government of Uganda.

UNCTAD (United Nations Conference on Trade and Development). Various years. Handbookof International Trade and Development Statistics. New York: United Nations.

Van Arkadie, Brian, and Gilles Durufle. 1986. "Planning in Africa: An Agenda for Revival."The Hague, Netherlands: Institute of Social Studies Advisory Service. Processed.

va. der Hoeven, R., and J. Vandermoortele. 1987. Stabilization and Adjustment Policies andProgrammes: Kenya. Country Study 4. Helsinki: World Institute for DevelopmentEconomics Research of the United Nations University.

Wangwe, Samuel. 1987. "Impact of the IMF/World Bank Philosophy: The Case ofTanzania." In Kjell T. Havnevik, ed., The IMF and the World Bank in Africa:Conditionality, Impact, and Alternatives. Seminar Proceeding 18. Uppsala, Sweden:Scandinavian Institute of African Studies.

Wheeler, David. 1984. "Sources of Stagnation in Sub-Saharan Africa." World Development,vol. 12, no. 1.

Williamson, John, ed. 1982. IMF Conditionality. Paper presented at a conference sponsored bythe Institute for International Economics on IMF conditionality. Washington, D.C.:Institute for International Economics. March.

Wood, Adrian. 1988. Global Trends in Real Exchange Rates, 1960 to 1984. World BankDiscussion Paper 35. Washington, D.C.: World Bank.

World Bank. 1981. Accelerated Development in Sub-Saharan Africa: An Agenda for Action.Washington, D.C.: World Bank.

. 1984. Toward Sustained Development in Sub-Saharan Africa: A Joint Program ofAction. Washington, D.C.: World Bank.

_ . 1985. Sudan: Pricing Policies and Structural Balance. Washington, D.C.: WorldBank.

. 1988a. World Development Report 1988. New York: Oxford University Press.

. 1988b. World Debt Tables, 1988-89 Edition: External Debt of Developing Countries.vol. 1: Analysis and Summary Tables. Washington, D.C.: World Bank.

_ . 1989. Strengthening Efforts to Reduce Poverty. Development Committee Paper 19.Washington, D.C.: World Bank.

World Bank and United Nations Development Programme. 1989. Africa's Adjustment andGrowth in the 1980s. Washington, D.C.

Young, Crawford. 1982. Ideology and Development in Africa. New Haven: Yale UniversityPress.

. 1985. "Optimism on Zaire: Illusions or Reality?" CSIS Africa Notes, no. 50.Washington, D.C.: African Studies Program of the Georgetown University Center forStrategic and International Studies.

References Ill

Zulu, Justin, and Saleh M. Nsouli. 1985. Adjustment Programs in Africa: The RecentExperience. IMF Occasional Paper 34. Washington, D.C.: International MonetaryFund.

Zuckerman, Elaine. 1989. Adjustment Programs and Social Welfare. World Bank DiscussionPaper 44. Washington, D.C.: World Bank.

Books in the ED)1 Seminar Series were designed for use in EDI courses and seminars or haveemerged from the presentations and discussions that took place in connection with theseactivities. They discuss issues in economic development policy and lessons from experience ina way that can be understood without extensive background knowledge or technical expertiseThey will be of particular interest to readers concerned with public affairs.

WORLD BANK PUBLICATIONS OF RELATED INTEREST

Poverty, Adjustment, and Growth in Africa.Ismail Serageldin. In English and French.

Adjustment Programs and Social Welfare.Elaine Zuckerman. World Bank Discussion Paper 44.

Africa's Adjustment and Growth in the 1980s.The World Bank and the LuNI)t. In English and French.

African Economic and Financial Data.The World Bank and the UNDP.

Public Enterprise in Sub-Saharan Africa.John Nellis World Bank Discussion Paper 1

Tax Policy in Sub-Saharan Africa: A Framework for Analysis.Zmarak Shalizi and Lyn Squire. World Bank Policy & Research Series 2. -

Impasse in Zambia: The Economics and Polities of Reform.Ravi Gulhati. El[ Development Policy Case Series, Analytical Case Study 2.

Malawi: Promising Reforms, Bad Luck.Ravi Gulhati. EDi Development Policy Case Series, Analytical Case Study 3.

Successful Development in Africa: Case Studies of Projects, Programs, andPolicies. EDI Development Policy Case Series, Analytical Case Study 1.

The Political Economy of Reform in Sub-Saharan Africa.Ravi Gulhati. EDI Policy Seminar Report 8.

Industrial Adjustment in Sub-Saharan Africa.Gerald M. Meier and William F.-Steel, editors. Oxford University Press.

I_*

The Making of Economic Policy in Africa - World...

Documents

Transcript of The Making of Economic Policy in Africa - World...